Introduction to Food Biotechnology [First edition] 9781315275703, 1315275708, 9781351989435, 135198943X, 9781420058383, 142005838X, 9781466519428, 1466519428, 0849311527

Table of contents :
Content: THE SCOPE OF FOOD BIOTECHNOLOGY --
Overview --
What Is Biotechnology? --
Recombinant DNA Technology --
Microbial Biotechnology --
Diagnostic Biotechnology --
Controversial Aspects of Food Biotechnology --
Food Security --
Recommended Reading --
TOOLS OF THE TRADE --
The Heart of Biotechnology: Cell Biology --
Bacteria --
Fungi are also Useful and Varied --
Viruses: Useful Parasites --
DNA: The Heart of Biotechnology --
Recommended Reading --
GENE CLONING AND PRODUCTION OF RECOMBINANT PROTEINS --
What Is a Recombinant Protein? --
Why Bother Making Recombinant Proteins? --
How and Why Are Genes Cloned? (The Big Picture) --
Gene Cloning: The Detailed Picture --
The cDNA Alternative --
Polymerase Chain Reaction: A Revolution in Cloning --
Alternative Vectors: pUC --
Alternative Vectors: Phage Lambda --
Artificial Chromosomes --
Subcloning and Vector Diversity --
Recombinant Chymosin --
Recombinant Bovine Growth Hormone --
Recommended Reading --
PLANT BIOTECHNOLOGY --
Overview --
Plant Cell and Tissue Culture --
Transgenic Plants --
Applications of Transgenic Plants to Food Production --
Transgenic Food Plants under Development --
Alternative Agricultural Technologies --
Recommended Reading --
ANIMAL BIOTECHNOLOGY --
Overview --
Transgenic Fish --
Modified Milk Proteins --
The Search for Embryonic Stem Cells --
Transfer of Somatic Nuclei: An Alternative to the Use of Embryonic Stem Cells --
Recommended Reading --
DIAGNOSTIC SYSTEMS --
Why are Diagnostic Systems Needed? --
Diagnostic Biotechnology --
Recommended Reading --
CELL CULTURE AND FOOD --
Overview --
Brewing --
Dairy Biotechnology --
Amino Acids: Nutritional Boosts and Flavour Enhancers --
Microbial Enzymes --
Microbial Polysaccharides --
Citric Acid and Vitamin Production --
Development of Novel Microbial Products --
Recommended Reading --
INDUSTRIAL CELL CULTURE --
Scale-Up of Cell Culture --
Environmental Factors --
Types of Bioreactors --
Downstream Processing --
Recommended Reading --
ETHICS, SAFETY, AND REGULATION --
Overall Perspective --
Consumer Perspectives and Food Biotechnology --
Safety Assessment and Regulation of Transgenic Crops --
Assessment and Regulation of Diagnostic Tests --
Biotechnology and the Developing World --
Recommended Reading --
Index

Citation preview

INTRODUCTION to

FOOD BIOTECHNOLOGY

CRC Series in CONTEMPORARY FOOD SCIENCE Fergus M. Clydesdale, Series Editor University of Massachusetts, Amherst

Published Titles: America’s Foods Health Messages and Claims: Scientific, Regulatory, and Legal Issues James E. Tillotson New Food Product Development: From Concept to Marketplace Gordon W. Fuller Food Properties Handbook Shafiur Rahman Aseptic Processing and Packaging of Foods: Food Industry Perspectives Jarius David, V. R. Carlson, and Ralph Graves The Food Chemistry Laboratory: A Manual for Experimental Foods, Dietetics, and Food Scientists Connie Weaver Handbook of Food Spoilage Yeasts Tibor Deak and Larry R. Beauchat Food Emulsions: Principles, Practice, and Techniques David Julian McClements Getting the Most Out of Your Consultant: A Guide to Selection Through Implementation Gordon W. Fuller Antioxidant Status, Diet, Nutrition, and Health Andreas M. Papas Food Shelf Life Stability N.A. Michael Eskin and David S. Robinson Bread Staling Pavinee Chinachoti and Yael Vodovotz Interdisciplinary Food Safety Research Neal M. Hooker and Elsa A. Murano Automation for Food Engineering: Food Quality Quantization and Process Control Yanbo Huang, A. Dale Whittaker, and Ronald E. Lacey

CRC Series in CONTEMPORARY FOOD SCIENCE

INTRODUCTION to

FOOD BIOTECHNOLOGY Perry Johnson-Green Department of Biology Acadia University Wolfville, Nova Scotia Canada

CRC PR E S S Boca Raton London New York Washington, D.C.

Library of Congress Cataloging-in-Publication Data Johnson-Green, Perry. Introduction to food biotechnology / Perry Johnson-Green. p. cm. -- (Contemporary food science (series)) Includes bibliographical references and index. ISBN 0-8493-1152-7 (alk. paper) 1. Food--Biotechnology. I. Title. II. Series. TP248.65.F66 J64 2002 664--dc21 2002017484

This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming, and recording, or by any information storage or retrieval system, without prior permission in writing from the publisher. The consent of CRC Press LLC does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from CRC Press LLC for such copying. Direct all inquiries to CRC Press LLC, 2000 N.W. Corporate Blvd., Boca Raton, Florida 33431. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe.

Visit the CRC Press Web site at www.crcpress.com © 2002 by CRC Press LLC No claim to original U.S. Government works International Standard Book Number 0-8493-1152-7 Library of Congress Card Number 2002017484 Printed in the United States of America 3 4 5 6 7 8 9 0 Printed on acid-free paper

Preface Biotechnology is becoming increasingly important to food. In some industries (e.g., brewing), it is part of a process that has deep roots in human society, whereas many other applications of biotechnology are new to food production and processing systems. Food biotechnology is also new to consumers; its introduction sometimes leads to opposition from consumer groups and antibiotechnology activist groups. In some cases, opposition has been strong enough to influence government policy toward regulation of biotechnology. However, many aspects of food biotechnology are virtually invisible to the consumer. Microbial products are increasingly common ingredients in processed foods, and the diagnostic tools used by the food industry to maintain food safety often have a biotechnological component. Consumers are becoming more aware of nutraceuticals and functional foods, and have enthusiastically embraced this aspect of biotechnology. Food scientists, nutritionists, dietitians, and agricultural professionals must have a rich understanding of food biotechnology, because biotechnology has the potential to be used as a tool within each of these disciplines. For example, transgenic plant biotechnology can be used to modify food or to improve its performance as a component of a processed food. Plants can also be modified to have higher nutrient or vitamin contents, increased levels of health-promoting compounds, or decreased levels of toxins or allergens. Transgenic plant technology has already been used extensively to improve the efficiency of food production, and there will be more applications of this sort in the future. One of the main objectives of this book is to lay a solid foundation in all areas of food biotechnology that can also be used as a springboard to careers in biotechnology. Readers will acquire an understanding of the language used in biotechnology, as well as the biological and chemical concepts that are important in each field. One of the major themes is biological diversity — the fount of most biotechnological innovation. Biotechnologists need to appreciate how the natural world has provided important tools to enhance technology. Another theme is the frequent use of examples. Some examples are currently used in the food industry, whereas others are taken from the research literature. Food professionals also need to be aware of the controversial aspects of food biotechnology. The final chapter reviews ethical and regulatory issues, but an effort has been made to discuss them throughout the book. For example, Chapter 4 includes a discussion of the potential of transgenic plants to harm nontarget insects such as the monarch butterfly. Chapters 3 and 7 also have sections devoted to specific controversies in food biotechnology (bovine growth hormone and eosinophilia–myalgia syndrome, respectively).

Each chapter closes with a list of recommended reading. These are a mixture of general sources which provide a wide range of supporting material for topics covered in the chapter and, more specific, which support examples used in the book. The order of the lists corresponds to the sequence of topics in the chapter. This book has benefited greatly from interactions and feedback with students in Food 3413 over the years. I am also indebted to Sheila Potter for training in Corel Draw and Krista Patterson for administrative support. Finally, I thank Julia GreenJohnson for her continual encouragement and advice.

Author Perry Johnson-Green has taught a senior course in food biotechnology since 1995, as well as courses in food microbiology, sensory science, and human biology at Acadia University in Wolfville, Nova Scotia. He has been involved in a wide range of research, covering neuroscience, plant–microbe interactions, and the potential use of plant-derived antimicrobial compounds as food preservatives. Current research topics include interactions between probiotic yeast and mammalian cell function. A member of the Canadian Institute of Food Science and Technology, the Canadian Botanical Association, and the Institute of Food Technologists, he frequently participates in public discussions on consumer issues in food biotechnology.

Table of Contents Chapter 1

The Scope of Food Biotechnology......................................................1

I. Overview .........................................................................................................1 II. What Is Biotechnology?..................................................................................3 III. Recombinant DNA Technology......................................................................4 A. Gene Cloning.............................................................................................4 B. Transgenic Plants ......................................................................................6 C. Recombinant Microbes .............................................................................9 D. Transgenic Animals.................................................................................11 IV. Microbial Biotechnology ..............................................................................11 A. Perspectives .............................................................................................11 B. Traditional Microbial Biotechnology......................................................13 C. Modern Microbial Biotechnology...........................................................13 V. Diagnostic Biotechnology.............................................................................13 VI. Controversial Aspects of Food Biotechnology.............................................14 VII. Food Security ................................................................................................16 Recommended Reading ...........................................................................................17 Chapter 2

Tools of the Trade ..............................................................................19

I. The Heart of Biotechnology: Cell Biology ..................................................19 II. Bacteria..........................................................................................................19 A. Bacterial Growth .....................................................................................19 B. Physiological Diversity ...........................................................................21 C. Bacterial Genetics ...................................................................................23 III. Fungi Are Also Useful and Varied................................................................25 A. General Characteristics of Fungi ............................................................25 B. The Use of Fungi in Recombinant DNA Technology............................28 IV. Viruses: Useful Parasites...............................................................................30 A. The Nature of Viruses .............................................................................30 V. DNA: The Heart of Biotechnology ..............................................................33 A. DNA Structure.........................................................................................33 B. DNA Replication.....................................................................................35 C. Transcription of mRNA ..........................................................................36 D. Editing of RNA Transcripts in Eukaryotes.............................................37 E. Translation of mRNA into Protein..........................................................37 F. Posttranslational Processing of Polypeptides .........................................38 G. Relevance of DNA to Biotechnology .....................................................39

H. Working with DNA .................................................................................40 1. Purificiation of Nucleic Acids............................................................40 2. Gel Electrophoresis ............................................................................40 3. Blotting and Hybridization.................................................................41 4. DNA Sequencing................................................................................42 Recommended Reading ...........................................................................................44 Chapter 3

Gene Cloning and Production of Recombinant Proteins ..................45

I. II. III. IV.

What Is a Recombinant Protein? ..................................................................45 Why Bother Making Recombinant Proteins?...............................................45 How and Why Are Genes Cloned? (The Big Picture).................................47 Gene Cloning: The Detailed Picture ............................................................49 A. Restriction Enzymes................................................................................49 B. Plasmid Vectors .......................................................................................51 V. The cDNA Alternative ..................................................................................57 VI. Polymerase Chain Reaction: A Revolution in Cloning................................59 A. Overview..................................................................................................59 B. The Mechanism of Amplification ...........................................................59 C. PCR Is Not Perfect..................................................................................63 D. Variations on PCR...................................................................................64 VII. Alternative Vectors: pUC ..............................................................................65 VIII. Alternative Vectors: Phage Lambda..............................................................67 A. Lambda Cloning Vectors.........................................................................67 B. Cosmids ...................................................................................................69 IX. Artificial Chromosomes ................................................................................70 X. Subcloning and Vector Diversity ..................................................................72 XI. Recombinant Chymosin ................................................................................72 A. Chymosin and Cheese Making ...............................................................72 B. Inclusion Bodies......................................................................................74 C. Recombinant Production by Yeast ..........................................................75 XII. Recombinant Bovine Growth Hormone .......................................................76 Recommended Reading ...........................................................................................79 Chapter 4

Plant Biotechnology...........................................................................81

I. Overview .......................................................................................................81 II. Plant Cell and Tissue Culture .......................................................................83 A. Control of Plant Growth .........................................................................83 B. The in vitro Life Cycle ...........................................................................84 C. Micropropagation ....................................................................................86 D. Plant Cell Culture and Traditional Plant Breeding.................................86 1. Plant Breeding Basics ........................................................................86 2. Protoplast Fusion ................................................................................89 3. Somaclonal Variation: Problem or Opportunity?...............................90

III. Transgenic Plants ..........................................................................................91 A. Overview..................................................................................................91 B. The Process of Making Transgenic Plants .............................................94 C. Agrobacterium tumefaciens: A Natural DNA Vector .............................97 D. The Natural Life Cycle of Agrobacterium tumefaciens .........................98 E. The Use of A. tumefaciens to Create Transgenic Plants ......................100 F. Alternatives to A. tumefaciens ..............................................................102 IV. Applications of Transgenic Plants to Food Production .............................103 A. Transgenics Resistant to Insect Pests ...................................................103 B. Pathogen Resistance ..............................................................................107 C. Herbicide Resistance .............................................................................110 D. Delayed Ripening..................................................................................111 E. Golden Rice...........................................................................................115 V. Transgenic Food Plants under Development..............................................119 A. Agronomic Traits...................................................................................119 1. Disease Resistance............................................................................120 2. Stress Resistance ..............................................................................120 B. Storage Proteins.....................................................................................121 C. Antinutrients and Other Undesirable Compounds ...............................121 D. Proteins Important to Bread Making ....................................................124 E. Engineering Better Starch .....................................................................126 F. Alternative Sweeteners..........................................................................126 G. Increased Levels of Vitamins and Phytochemicals ..............................127 H. Reducing or Eliminating Allergens.......................................................129 VI. Alternative Agricultural Technologies ........................................................131 Recommended Reading .........................................................................................133 Chapter 5

Animal Biotechnology .....................................................................137

I. II. III. IV. V.

Overview .....................................................................................................137 Transgenic Fish ...........................................................................................139 Modified Milk Proteins...............................................................................142 The Search for Embryonic Stem Cells.......................................................145 Transfer of Somatic Nuclei: An Alternative to the Use of Embryonic Stem Cells ...................................................................................................148 Recommended Reading .........................................................................................149 Chapter 6

Diagnostic Systems ..........................................................................151

I. Why are Diagnostic Systems Needed?.......................................................151 A. Overview and Global Perspective .........................................................151 B. Diagnostics and Hazard Analysis Critical Control Points (HACCP) ...............................................................................................153 C. Nonpathogen Diagnostics .....................................................................156

II. Diagnostic Biotechnology...........................................................................158 A. Scope .....................................................................................................158 B. Nucleic Acid Probes..............................................................................159 C. Exploiting the Polymerase Chain Reaction ..........................................162 D. DNA Chips and Micro-Arrays..............................................................163 E. Antibody-Based Diagnostic Systems....................................................164 1. Applications of Antibodies to Diagnostics ......................................164 2. Manipulating the Immune Response ...............................................168 3. Future Applications of Immuno-Assays ..........................................171 F. Luminescence and Diagnostics.............................................................173 1. Hygiene Assessment.........................................................................173 2. Novel Applications of Luminescence ..............................................173 G. Biosensors..............................................................................................174 1. Applications of Biosensors ..............................................................174 2. Types of Biosensors .........................................................................176 Recommended Reading .........................................................................................179 Chapter 7

Cell Culture and Food......................................................................181

I. Overview .....................................................................................................181 II. Brewing .......................................................................................................182 A. Introduction ...........................................................................................182 B. Malting ..................................................................................................185 C. Mashing .................................................................................................187 D. Hops.......................................................................................................188 E. Primary Fermentation............................................................................189 F. Secondary Fermentation........................................................................190 G. Biotechnological Improvements: Catabolite Repression......................191 H. High-Gravity Brewing...........................................................................192 I. The β-Glucan Problem..........................................................................194 J. Getting Rid of Diacetyl.........................................................................194 III. Dairy Biotechnology ...................................................................................195 A. Introduction ...........................................................................................195 B. Starter Cultures......................................................................................197 C. Phage .....................................................................................................198 D. Recombinant Lactic Acid Bacteria .......................................................199 1. Improved Starters .............................................................................199 2. Recombinant Lactic Acid Bacteria as Vaccines...............................201 IV. Amino Acids: Nutritional Boosts and Flavor Enhancers ...........................202 A. Overview................................................................................................202 B. Choice of Microbe ................................................................................204 C. Proline: Mutagenesis Leads to Overproduction ...................................208 D. Glutamate: Natural Overproduction......................................................209 E. Aspartate: Conversion by Immobilized Enzymes ................................211 F. Tryptophan and Eosinophilia–Myalgia Syndrome ...............................212

V. Microbial Enzymes .....................................................................................213 A. Overview................................................................................................213 B. Amylases ...............................................................................................215 1. Starch Processing..............................................................................215 2. Amylases and Baking.......................................................................218 C. Lipases...................................................................................................219 D. Polygalacturonase..................................................................................220 VI. Microbial Polysaccharides ..........................................................................221 A. Overview................................................................................................221 B. Complex Polysaccharides .....................................................................222 C. Xanthan Gum ........................................................................................224 1. Structure and Characteristics............................................................224 2. Xanthomonas campestris..................................................................225 3. Genetics of Xanthan Gum Biosynthesis ..........................................225 VII. Citric Acid and Vitamin Production ...........................................................226 A. Uses of Citric Acid in Food..................................................................226 B. Production of Citric Acid by Aspergillus niger....................................227 C. Vitamin Production by Microorganisms...............................................228 VIII. Development of Novel Microbial Products................................................229 Recommended Reading .........................................................................................231 Chapter 8

Industrial Cell Culture .....................................................................233

I. Scale-Up of Cell Culture ............................................................................233 II. Environmental Factors ................................................................................235 A. Oxygen ..................................................................................................235 B. pH ..........................................................................................................240 C. Temperature ...........................................................................................241 D. Nutrient Supply .....................................................................................243 III. Types of Bioreactors ...................................................................................246 A. Stirred-Tank Bioreactors .......................................................................246 B. Continuous Culture ...............................................................................249 1. Advantages of Continuous Culture ..................................................249 2. Disadvantages of Continuous Culture..............................................251 3. Continuous Culture in Operation .....................................................251 C. Immobilized Cells and Enzymes ..........................................................253 1. Immobilized Cells ............................................................................253 2. Immobilized Enzymes ......................................................................257 IV. Downstream Processing ..............................................................................258 A. The Importance of Downstream Processing.........................................258 B. Cell Lysis...............................................................................................259 C. Separating Solids from Liquids ............................................................259 D. Phase Changes.......................................................................................261 Recommended Reading .........................................................................................262

Chapter 9

Ethics, Safety, and Regulation .........................................................265

I. Overall Perspective .....................................................................................265 I. Consumer Perspectives and Food Biotechnology ......................................267 III. Safety Assessment and Regulation of Transgenic Crops...........................269 A. Assessment Strategies ...........................................................................269 B. The Substantial Equivalence Debate ....................................................271 C. Risk Assessment of Novel Genetic Elements.......................................272 D. Starlink Corn and Implications for Labeling and Trade Issues ...........275 IV. Assessment and Regulation of Diagnostic Tests........................................276 V. Biotechnology and the Developing World .................................................276 VI. The Future of Food Biotechnology ............................................................278 Recommended Reading .........................................................................................278 Index......................................................................................................................281

1

The Scope of Food Biotechnology I. OVERVIEW

Food is central to human society. Food production and processing are vital components of global and local economies, and all aspects of the food industry rely heavily on technology. The development of food technology has been one of the great success stories of the 20th century. In the western world, we enjoy the luxury of year-round access to a dizzying diversity of fresh and processed foods, due to technological improvements in our ability to grow, store, and process plant and animal foods. In developing countries, food production has generally increased as populations increase, largely because of the development of high-yielding seeds. However, from a global perspective, it is easy to find fault with current food technology. In industrialized countries, food production relies heavily on fossil fuels and products derived from them, such as synthetic pesticides and chemical fertilizers (production of fertilizers requires large inputs of energy, usually derived from fossil fuels). Hunger and malnutrition are still common in the developing world, despite attempts to increase the efficiency of food production in those regions. Throughout the world, arable land is becoming scarce; erosion and salinization are placing greater stresses on crops. The human population also continues to grow. Although the pace of growth has slowed considerably in the last 20 years, most forecasts call for continued population growth, particularly in the developing world. Increasing food production to accommodate for this growth is a challenge. Many problems are felt on a smaller scale, in relation to the conversion of food commodities into valuable products. The incidence of foodborne illness is increasing in many parts of the world, and this increase has partly been blamed on more complex food processing systems. Increased consumer demand for relatively unprocessed foods has also contributed to the risk of food-borne illness. Recent cases of Escherichia coli O157:H7 infection from unpasteurized apple cider are an example of this phenomenon. Consumers are more aware of the dangers of food-borne pathogens and are more concerned about the potential health risks associated with pesticide residues on food. Biotechnology certainly has the potential to alleviate many of these problems. Such efforts may lead to more efficient and sustainable agricultural systems and reduced reliance on chemical pesticides and fertilizers. One objective of this book is to develop an understanding of the approaches that can be taken to improve food production systems. For example, the application of recombinant DNA technology to directly modify an organism’s genetic structure has provided producers of corn, cotton, and potato an alternative to synthetic pesticides to control insect pests. The

1

2

Introduction to Food Biotechnology

ability to transfer DNA from a bacterium (Bacillus thuringiensis) to a plant, a process that was unthinkable 40 years ago, is crucial to the development of these improved plants. B. thuringiensis produces toxic proteins that kill a narrow range of insects. When the gene for this type of toxic protein is transferred to a plant, thus making a transgenic plant, the plant becomes toxic to susceptible insects. This technology can also be used to improve the nutritional qualities of plant foods and improve processing characteristics (e.g., modifying the characteristics of high-molecular glutenins in wheat to increase the rising power of bread doughs). Biotechnology also has the potential to improve food safety, through the development of enhanced diagnostic systems to detect pathogens and toxins in food. Microbes also play a positive role in the food industry; -microbial biotechnology is an increasingly important part of food processes. A variety of valuable enzymes, amino acids, and polysaccharides can be obtained from bacteria and fungi, and their use is steadily increasing in the food industry. Why, then, is food biotechnology so controversial? Consumers, particularly in Europe, Britain, and Japan, are particularly hostile to transgenic crops. Many consumer and environmental groups believe that these crops are potentially dangerous to humans and the environment and that these risks have been insufficiently assessed. There is also a common perception that food biotechnology is beneficial only to the companies involved in its development, and not to the general public. It is also a favorite target of antiglobalization activists, who view biotechnology as a means to exploit producers in the developing world. Public concern about biotechnology increased in the fall of 1999, when largescale contamination of “Starlink” corn was found in taco shells and various other foodstuffs destined for human consumption. This variety of corn contains a gene that produces a protein toxic to the European corn borer, a destructive pest. However, it was never approved for human consumption; prior to the Starlink scandal, however, it could be used in animal feed. The reason for this regulatory strategy was a concern that the novel protein had the potential to provoke an allergic response in humans, because it was somewhat resistant to breakdown by digestive enzymes. As a result of this contamination, several companies suffered substantial fiscal damage, and public distrust of food biotechnology grew. Soon after the Starlink scandal, Swiss biotechnologists announced the development of “golden rice,” a variety of rice that has much higher levels of β-carotene than normal rice. The human body can use β-carotene to synthesize vitamin A. Vitamin A deficiency is widespread in the developing world and is the leading cause of noninfectious blindness. The most common reason for this deficiency is overreliance on rice, which is often the only food available to people living in dire poverty and to subsistence farmers in southeast Asia. It is uncertain how effective golden rice will be in combating vitamin A deficiency, and it is not a permanent solution to global malnutrition, but it will probably lead to improvements in vitamin A nutrition of some of the world’s poor. The biotechnology industry has been quick to capitalize on golden rice, using it as an example of the potential of biotechnology to beneficially affect human society. This use of golden rice as a public relations tool has been widely criticized, especially

The Scope of Food Biotechnology

3

by the developers of golden rice. The pathway of development of golden rice has been very different from that of commercial transgenics. Funding for development came from the Swiss government and an American foundation (the Rockefeller Foundation). The developers of golden rice have always been adamant that golden rice seed would be freely available to farmers in the developing world and that they would not seek intellectual property rights for this transgenic plant. In contrast, companies such as Monsanto have successfully obtained patents for their transgenic crops, and patent rights are vigorously defended. A number of high-profile court cases in the U.S. and Canada have demonstrated that the biotechnology industry considers intellectual property of transgenic crops to be essential. Consumers throughout the world, then, are faced with contradictory information about biotechnology. Numerous surveys have also found that consumers tend to misunderstand the nature of biotechnology. For example, a survey of Australian consumers in 2000 found that most do not have a clear understanding of the potential risks and benefits of food biotechnology, yet they feel that the risks are greater than the benefits. Thus, consumers are suspicious of the entrance of biotechnology into the food industry, but this suspicion is not driven by specific knowledge. Consumers in the rest of the world are likely in a similar state. This is a crucial time for food biotechnologists; the public is generally hostile to biotechnology, but there seems to be room for improvement in attitudes. To encourage this change, the biotechnology industry needs to show consumers that biotechnology has the potential to improve their lives, without undue risk to the environment or to human health. Most consumers are unaware of the less controversial aspects of food biotechnology, such as microbe-derived food additives (e.g., xanthan gum) and DNA-based diagnostic techniques. These types of biotechnology offer few potential risks to consumers and the environment and have generally been ignored by antibiotechnology activist groups. The main objective of this book is to achieve an understanding of all aspects of food biotechnology. This will enable the reader to focus on specific technologies with more advanced readings, as needed. Risks and benefits will also be explained, with reference to numerous examples, allowing a thorough understanding of the controversies that surround food biotechnology.

II. WHAT IS BIOTECHNOLOGY? Because of the wide range of processes that have been described as biotechnology, it is difficult to concisely define the term. However, as applied to food, biotechnology usually describes one of the following processes: • The direct modification of DNA of plants, animals, or microbes that are used as food (popularly known as “genetically modified organisms,” or GMOs) • The use of microbes or microbial products as food or food additives • DNA- or protein-based methods for the detection or identification of microbes or microbial products in food

4

Introduction to Food Biotechnology

This list is not exclusive; many aspects of traditional plant breeding include processes that are commonly considered to be biotechnology. For example, in their search for new useful traits, plant breeders often attempt to cross crop plants with their wild relatives. This is usually facilitated by various techniques of cell culture, which is generally considered to be a biotechnological process. Directed use of microbes in agriculture is also becoming more common. Rhizobium, a bacterium that is able to transfer atmospheric nitrogen to certain plants, thus decreasing reliance on fertilizers, is commonly inoculated onto seeds. Microbial biocontrol of insect and weed pests is also increasingly popular. Functional foods and nutraceuticals are sometimes considered to be part of biotechnology. Functional foods are usually defined as foods that are part of a normal human diet that provide benefits to health beyond the supply of basic nutrients. Nutraceuticals are sometimes defined as the specific compounds within functional foods that are responsible for their health-promoting effects; more commonly, though, the term “nutraceutical” is restricted to the marketing of food components that improve health and are sold in a purified form (e.g., as a tablet). Because functional foods are usually derived from common, unmodified plants produced through conventional agriculture, they do not fit well with other forms of food biotechnology. For this reason, they will not be discussed extensively in this book. However, the use of recombinant DNA technology to increase levels of vitamins is clearly biotechnology, but also has functional food implications, because many vitamins have health-promoting effects (e.g., as antioxidants) that are distinct from their traditional role as vitamins. Many researchers in the biotechnology community are very interested in the potential of recombinant DNA technology to increase levels of other food components that improve health or prevent disease.

III. RECOMBINANT DNA TECHNOLOGY A. GENE CLONING The major breakthrough in the development of recombinant DNA technology was the ability to clone genes. This refers to the process of isolating a specific gene from an organism’s genome (the entire set of genetic information in an organism). In Chapter 3, we will discuss in detail how this is done, but in general terms, genes are usually cloned by inserting fragments of a genome into a vector (Figure 1.1). A vector is an agent that can be used to move DNA segments from one organism to another. Plasmids, small circular double-stranded DNA molecules that are capable of replication within their host cell, are commonly used as vectors. Once a plasmid vector has been inserted into a cell, the cell that contains the desired gene can be located and separated from cells that contain other fragments of DNA. Gene cloning allows careful study of a gene’s sequence and properties, and it also allows the gene to be transferred to a wide variety of organisms. Thus, a gene isolated from a bacterium can be transferred to another bacterium, a plant, or an animal. In some cases, gene transfer is relatively easy; in others (e.g., inserting a gene into a multicellular animal), it is much more challenging and complex. A defining feature of molecular biotechnology is the ability to transfer specific genes

The Scope of Food Biotechnology

5

Find organism with the desired gene

Cut genome into fragments

Insert fragments into vector

Plasmid vector

Insert vector into host cell

Isolate cell with the desired gene

*

FIGURE 1.1 Overall process of gene cloning. In this example, a desirable gene is present in a carrot cultivar. DNA from the carrot is cut into small fragments and inserted into a plasmid vector, which is then inserted into a host cell. The cell with the desired gene is then located and isolated.

from organism to organism without the restrictions of incompatibility that otherwise apply (e.g., animals will breed successfully only with animals of the same species). The basic techniques of gene cloning were developed in the mid-1970s. A product of direct gene transfer is considered to be recombinant, because its genome now consists of DNA from different organisms. The transfer process is known as genetic engineering, and in the popular media, the products are known as genetically modified organisms (GMOs). In this book “GMOs” will be used in general reference to organisms (plants, animals, and microbes) that have been modified using recombinant DNA technology, and the terms “transgenic plants,” “transgenic animals,” and “recombinant microbes” will be used when referring to specific types of GMOs.

6

Introduction to Food Biotechnology

B. TRANSGENIC PLANTS Sometime after the last glaciation (~11,000 years ago), humans began to cultivate plants and herd animals for food. They probably also began to breed these crops and animals. Breeding is fundamentally a simple process: parents with superior traits are allowed to mate (for plants, this means that pollen from one parent is used to pollinate flowers of the other parent), and offspring that have desirable traits are then selected and allowed to breed. This process continues until the desired improvement is obtained. Breeding is still an effective mechanism for improvement and is frequently used to increase such traits as yield and disease resistance, as well as characteristics important to food processors (e.g., sucrose levels in potatoes) and nutritionists (e.g., levels of β-carotene in carrots). However, the great disadvantage that breeders face is the lack of control over the gene mixing that occurs during normal sexual reproduction. When a nucleus from a pollen grain fertilizes a nucleus in an ovule all of the chromosomes of the pollen nucleus are mixed with all of the chromosomes of the egg cell in the ovule. In many cases undesirable traits are passed to the egg cell along with desirable traits. In the mid-1970s, plant scientists were quick to see the potential of recombinant DNA technology to revolutionize plant breeding. Instead of being dependent on characteristics found in other cultivars, or closely related species, plant breeders could transfer genes from virtually any other organism. Furthermore, the wholesale mixing of genomes that occurs during conventional (traditional) breeding, does not occur with recombinant DNA technology. Only the desired gene and one or two other genes, depending on the methodology used, are transferred to the recipient plant. This means it is less likely that valuable traits will be lost during the gene transfer. In the early 1980s, the first studies of transgenic plants were published, and in the early 1990s, several companies took initial steps toward commercialization of transgenic plants. Transgenic crop plants have received intense public scrutiny and remain highly controversial. Since 1994, 51 transgenic crop varieties have been released into the U.S. (Figure 1.2). Most of these were released between 1995 and 1998, and most confer benefits to producers; they provide herbicide resistance, contain genes that produce proteins that are toxic to insect pests, or give resistance to viruses that are plant pathogens (Table 1.1). Relatively few of the released crops are directly relevant to consumers, except for varieties with delayed ripening. Delayed-ripening tomatoes offer two potential advantages: a longer period of ripening in the field, without the softening that normally accompanies vine ripening, and an increased content of solids, due to the decreased rate of pectin breakdown. The first characteristic allows better transport of vine-ripened tomatoes, and the second allows the production of improved tomato paste. In the future, there will likely be an increase in development and release of transgenic plants that have direct benefits to the food industry and to consumers (Table 1.2). Currently, few such transgenics exist. Why? One reason is that modifications aimed at processing or consumer problems often are technically difficult to achieve. For example, the suppression of lipoxygenase (LOX) activity in peas and other legumes improves flavor and aroma. However, it may also lead to decreased

The Scope of Food Biotechnology

7

Number of releases

16 12 8 4 0 1994

1995

1996

1997

1998

1999

2000

Year

FIGURE 1.2 Commercial releases of trangenic crops in the U.S. between 1994 and 2000.

TABLE 1.1 Introduced Traits in Recombinant Crops Released in the U.S. between 1994 and 2000 Trait Herbicide resistance Insect resistance Sterility/fertility Delayed ripening Virus resistance Modified lipid profile Improved nutritional content

Number of Releases 23 14 8 6 3 2 1

Note: Some crop varieties contain more than one introduced trait.

plant resistance against various stresses, such as insect attack. Therefore, such modification must be done with a thorough understanding of its implications. Few of the world’s crops have been modified through recombinant DNA technology (Table 1.3). Rice and wheat have been particularly difficult to transform, partly because of difficulties in growing rice and wheat cells in vitro. However, recent advances have made the transformation of wheat and rice relatively straightforward; the number of recombinant wheat and rice releases is likely to increase in the next five years. Commercial development of transgenics of minor crops (e.g., most vegetable crops) is unlikely until there is greater public acceptance of transgenic plants. Recombinant crops have proven to be very popular with producers. Every year, the Economic Resource Service (ERS) of the U.S. Department of Agriculture

8

Introduction to Food Biotechnology

TABLE 1.2 Potential Uses of Transgenic Plants Relevant to Consumers and the Food Industry Plant

Modification

Tomato and other plants Tomato Corn Corn, canola, etc. Various plants

Delayed ripening Increased chitinase Control over starch structure Control over lipid profile Addition of phytase

Advantage

Easier transport of fruits; improved quality Less post-harvest spoilage Fewer requirements for starch conversiona Oils that promote human health Decreased antinutritional compounds (phytate) Legumes Suppression of protease inhibitors Increased digestibility Soybean Suppression of lipoxygenase Improved flavor Improved bread quality Wheat Increased HMWb glutenin Barley Increased β-glucanase levels Fewer haze problems in beer Various plants Modification of enzyme activity Increased antioxidants Peanut and other plants Elimination of allergens Less allergenicity Rice, tomato, etc. Increased provitamin A Increased vitamin A supplyc a

Starch is normally converted to a range of products (e.g., maltodextrins) through the use of microbial enzymes. b High molecular weight. c Transgenics with increased vitamin E have also been developed.

TABLE 1.3 Recombinant Crops Released in the U.S. between 1994 and 2000 Crop Corn Canola Tomato Cotton Potato Soybeans Squash Cantaloupe, flax, papaya, rice

Number of Releases 13 10 6 5 4 3 2 1 each

(USDA) surveys a random selection of farmers. These surveys indicate progressive increases in acreage of recombinant crops between 1996 and 2000, with some recombinant varieties achieving close to 50% of total acreage (e.g., herbicideresistant soybeans). Insect-resistant and herbicide-resistant crops have been particularly successful, in some cases because of decreased costs of chemical control of insect or weed pests.

The Scope of Food Biotechnology

9

Such large plantings means that much of the corn, soybean, and canola used in food processing in North America are transgenic. However, because only a few transgenic crops (one cultivar of insect-resistant corn and one cultivar of herbicideresistant soybeans) can currently be imported and sold in food in Europe, some of the North American–grown transgenic cultivars must be separated. This process is known as crop segregation, and for some commodities (e.g., soybeans in the U.S.), it is difficult because such segregation was not necessary before the introduction of transgenic crops. Thus, systems for separate handling of different cultivars have to be introduced into current systems, which may be an expensive process. Many farmers, particularly after the Starlink scandal, fear that such requirements for segregation may impede international trade in transgenic crops. Consequently, they must balance any benefits of growing recombinant crops against potential difficulties in selling their crop. Problems related to contamination of GMO-free crops by transgenic seed have also occurred in the last 5 years. This, along with the Starlink scandal, demonstrates that it is technically difficult to completely segregate commodities based on their recombinant status. The contamination problem is likely to hinder commodity exports in the future, unless global consumer distrust of recombinant crops decreases in intensity. Recently released (July 2001) proposed regulations governing the use of transgenic crops in food in the European Union (EU) state that any food containing more than 1% transgenic products must be labeled to indicate that it contains GMOs. Thus the presence of low levels of transgenic seed in bulk commodity shipments of non-GMO seed could cause problems. In Europe, no new transgenic plants will be approved until 2002. However, the process for approval of new transgenic cultivars is still under development, and the EU is not expected to implement the new process until 2003. Therefore, it is unlikely that large increases in the sowing of transgenic crops in Europe will occur in the near future. Anti-biotechnology activist groups are much more vocal and successful in Europe than in North America, and the European public remains unconvinced of the human and environmental safety of recombinant foods. Attitudes toward transgenic crops appear to be improving in Britain, although the widespread antagonism toward experimental plots shows that the atmosphere is still not encouraging for producers who wish to grow transgenic crops.

C. RECOMBINANT MICROBES To date, the use of recombinant microbes in food production and processing has been limited to recombinant microbial enzymes and a recombinant hormone (bovine growth hormone [BGH]) to boost milk production. Recombinant chymosin (rennet) is an example of a recombinant enzyme produced by microbes. The bovine chymosin gene was transferred to several fungal species via recombinant DNA technology in the 1980s. Recombinant chymosin is now widely used throughout the world in cheese making. Chymosin increases the rate of curd formation during initial fermentation of milk by lactic acid bacteria (Figure 1.3). Traditionally, chymosin was obtained from the stomach of slaughtered calves, but the supply from this source is somewhat unstable. In contrast, recombinant chymosin does not have this instability, because it can be produced through growth of recombinant yeasts in large bioreactors

10

Introduction to Food Biotechnology

Pasteurize milk

Add chymosin

Milk acidifies and coagulates

Compress the solid curd

Remove the liquid whey

Add salt and/or ripen

FIGURE 1.3 The process of cheese making, illustrating the importance of chymosin.

(vessels used for large-scale growth of cells). Interestingly, recombinant chymosin is currently triggering hostility between the U.S. and the EU. Proposed EU regulations governing labeling of food containing GMOs exempt cheese made using recombinant chymosin, because such cheese, when it is sold to consumers, contains negligible amounts of recombinant chymosin. However, the U.S. government maintains that this constitutes an unfair trade practice, because it implies that European products such as cheese are exempt from labeling, whereas foods containing small amounts (>1%) of transgenic plants (mostly imported from the U.S.) must be labeled. The other major use of recombinant microbes in food is recombinant bovine growth hormone (rBGH) by dairy farmers. The gene for BGH was transferred to E. coli, and large-scale culture of this recombinant bacterium yields large amounts of rBGH. This is then injected into cows, which increases milk production. The American public largely opposed the introduction of rBGH into the dairy industry (see Chapter 3) even though it was deemed to be safe by the U.S. Food and Drug Administration (FDA). It cannot be used legally in Canada, Europe, and many other countries. In some cases, this is due to animal welfare concerns, and in others, concern over potential effects on human health. It is worth noting that most of the animal and health concerns are linked to the hormone itself and its effect on bovine metabolism rather than to the fact that rBGH is recombinant. In other words, injection of natural BGH into cows would lead to similar safety concerns.

The Scope of Food Biotechnology

11

D. TRANSGENIC ANIMALS Given the commercial success of transgenic plants, the lack of commercially available transgenic animals is surprising. There are several reasons for this. Many crops had readily identifiable problems (e.g., corn and the European corn borer) that could be attacked with single genes (e.g., a gene that produces a protein that is toxic to the borer). Such simple problems are less common in animal production systems. For example, feed efficiency (the ability of an animal to convert feed into tissue) is tremendously important; if an animal can gain weight with less feed, the producer can reap large savings. Unfortunately, a number of factors control feed efficiency, and single-gene solutions are unlikely. Increased animal (e.g., pig) growth hormone improves feed efficiency but can also seriously affect animal health (e.g., abnormal skeletal development). Fish are an exception. Transgenic salmon with boosted levels of growth hormone grow substantially faster than their nonrecombinant counterparts, apparently without major side effects. However, environmental concerns are substantial; transgenic fish would be raised in aquaculture, and in such systems it is difficult to prevent occasional escapes into the environment. Escaped fish could then interbreed with wild salmon populations, possibly resulting in decreased levels of overall fitness (see Chapter 5). Transgenic fish are usually sterile; thus escaping fish would be unable to breed with wild fish. Despite this protective feature, the release of transgenic fish remains a contentious issue. The highly publicized success of Scottish researchers in producing an adult clone of a sheep (“Dolly”) has focused attention on the possibility of using transgenic animals to produce human proteins in sheep. Such proteins (e.g., α1-antitrypsin) have therapeutic uses. This “molecular pharming” will probably be the first commercial use of transgenic farm animals; it is debatable whether such use of animals is either a “food” or “agriculture.” Nonetheless, the success or failure of molecular pharming in animals will have important reverberations on the development of transgenic animals with more conventional agricultural uses. Such uses include the modification of milk proteins or butterfat levels, the enhancement of resistance to viral or bacterial pathogens, and increased leanness of meat. Commercialization of transgenic animals is unlikely to happen before 2003, when the first transgenic salmon are expected to be ready for commercial use and sale, if regulatory approval is obtained. Many animal reproductive technologies also fall under the rubric of biotechnology. Embryo cloning, preservation, and transfer are all part of modern animal husbandry and animal breeding programs. If an individual animal has particularly good characteristics, its offspring can be multiplied extensively using in vitro (cell culture) techniques, allowing the widespread dissemination of animals with good traits.

IV. MICROBIAL BIOTECHNOLOGY A. PERSPECTIVES Micro-organisms are extremely important in biotechnology. Bacteria and fungi can be grown in large scale using bioreactors, which are large vessels that typically allow

12

Introduction to Food Biotechnology a

b

c

Prepare feed

Gelatinize starch

Pasteurize milk

Inoculate with Xanthomonas campestris

Add microbial amylases

Inoculate milk with lactic acid bacteria

Purify xanthan gum

Convert starch to various products (e.g., glucose)

Package into containers

Use as food ingredient (e.g., as a thickener)

Use as food ingredients (e.g., as a sweetener)

Incubate at 43°C

Cool to 4°C

FIGURE 1.4 Examples of food processes that rely on microbes. (a) Production of xanthan gum by the bacterium Xanthomonas campestris. (b) Use of microbial enzymes (amylases) to convert starch into useful food ingredients. (c) Use of lactic acid bacteria to convert milk to yogurt.

close control of nutrient and oxygen levels, pH, and other environmental factors. In relation to the food industry, microbes are grown in bioreactors for one of four reasons: (1) they produce a compound that is useful, either as a food additive or as a food; (2) they produce enzymes that can be used to modify the properties of foods or food ingredients; (3) they transform a food into a different type of food; or (4) they transform waste products of the food industry (e.g., cellulose) into less environmentally harmful products (e.g., CO2). Xanthan gum is an example of a microbederived food ingredient; this polysaccharide is a useful thickening agent used in salad dressings and other processed foods (Figure 1.4a). It is produced by Xanthomonas campestris, a relatively common bacterium. Modified starches are another group of common food ingredients. They are produced through the use of amylases and other microbial enzymes to alter the structure of corn or potato starch (Figure 1.4b). Common sweeteners such as glucose and fructose as well as various thickeners and fillers are obtained in this way. Yogurt is an example of a food that is made through microbial transformation of one food (milk) into another food (yogurt; Figure 1.4c). The brewing, wine-making, and distillery industries rely on the ability of the yeast Saccharomyces cereviseae to convert carbohydrates to ethanol. We can make a distinction between traditional and modern microbial biotechnologies. Dairy and ethanol fermentations are the most common traditional biotechnologies; they have been practiced for millennia and are well accepted by consumers

The Scope of Food Biotechnology

13

in most parts of the world. In contrast, xanthan gum and other microbial food additives and enzymes were not part of the food industry until the 20th century.

B. TRADITIONAL MICROBIAL BIOTECHNOLOGY All alcoholic beverages are produced using traditional biotechnological processes. Yeasts are added to carbohydrate substrates such as sucrose; the fungi use these substrates as a source of carbon and energy, and ferment them into ethanol and carbon dioxide. Yeasts are essential to the process. Without yeast, ethanol is not produced, and many of the flavors characteristic of each beverage are absent. Yeasts are also used in bread making. The production of carbon dioxide by yeast results in the formation of gas pockets, which drive the rising process. Cheese, yogurt, and other fermented milk products are also made with the help of microbes (see Figures 1.3 and 1.4). In this case the lactic acid bacteria are vital. This group of bacteria produces energy through fermentation, and one of the main products of fermentation (often the sole product of fermentation) is lactic acid. This organic acid decreases the pH of milk, which causes thickening and coagulation of milk proteins, and creates an environment antagonistic to bacteria that are pathogens (e.g., Salmonella spp.) or that produce off flavors in milk (e.g., Pseudomonas spp.). The lactic acid bacteria (e.g., Lactobacillus bulgaricus) also produce volatile compounds that contribute to the flavor of fermented milk products.

C. MODERN MICROBIAL BIOTECHNOLOGY Alcohol and lactic acid are not the only microbial products used in the food industry. Many of the enzymes, amino acids, and thickeners that are added to food are derived from a variety of microbes. In addition to xanthan gum, glutamate is also a microbial product that is a common food ingredient. It is added to many foods (e.g., dehydrated soups) to add “brothy” flavor and enhance other flavors. Glutamate is an amino acid and is therefore found in most proteins. However, it is easier to collect it from microbes than to separate glutamate in protein (e.g., soy protein) from other amino acids. Corynebacterium glutamicum, a Gram-positive bacterium, naturally has the ability to secrete large amounts of glutamate. This ability has been exploited by biotechnology companies, particularly in Japan.

V. DIAGNOSTIC BIOTECHNOLOGY Food safety is an essential element of food security. In industrialized nations, food safety is currently of prime concern to the public and is vital to commercial success in the food industry. Pathogens such as E. coli O157:H7, Campylobacter jejuni, Listeria monocytogenes, and Salmonella enteritidis are increasingly frequent causes of outbreaks of food-borne illness, leading to great economic costs, and, in some cases, death. Overall rates of food-borne illness are also increasing, and combined with scandals such as the bovine spongiform encephalopathy (BSE) epidemic in the U.K., this has led to profound and widespread unease among consumers and governments and throughout the food industry.

14

Introduction to Food Biotechnology

The ability to detect and correctly identify contaminants in food is a vital part of the ongoing fight against food-borne illness. Traditional diagnostic methods for bacteria are effective but often time consuming. This limits their usefulness, because only limited numbers of samples can be taken, and often, contamination is undetected until illness occurs. Many key diagnostic techniques have a biotechnological component. The most common of these is the use of mammalian antibodies to confirm the identification of bacterial species and to determine strain identity of bacteria. Such antibodies are produced through manipulation of the immune response in mammals such as mice and rabbits. They can also be obtained from hybridoma cells grown in vitro; (this term refers to the growth of cells outside of their normal location within a multicellular organism). Antibodies derived in this way are referred to as monoclonal antibodies. Antibody-based systems are also widely used clinically to identify viruses from samples of body fluids of people with food-borne illnesses. DNA-based methods are also starting to make an impression in clinical and food-industry settings. The polymerase chain reaction (PCR) and the use of labeled probes can potentially increase the speed and sensitivity of methods in detecting and identifying pathogens. PCR is very sensitive; theoretically, the DNA of a single pathogen could be amplified and detected using PCR. One of the major problems with DNA-based methods is the complexity of foods; many compounds in most foods interfere with these methods. Nevertheless, we expect DNA-based diagnostics will be increasingly important in the food industry. At the present time, these methods are mainly used for the identification of particular strains of pathogens and the detection of transgenic plant DNA in commodities such as soybeans and corn.

VI. CONTROVERSIAL ASPECTS OF FOOD BIOTECHNOLOGY Some aspects of food biotechnology are relatively free from controversy. Few consumers are aware of the use of microbial products such as recombinant chymosin in cheese or xanthan gum in salad dressings. The production of ethanol by yeasts is considered to be a safe technology, although ethanol abuse remains a troublesome issue for societies throughout the world. Diagnostic biotechnologies are seen as beneficial and powerful tools that can potentially decrease the incidence of foodand water-borne illness. Transgenic plants and animals, however, have aroused the attention of activist and consumer groups, particularly in Britain and Europe. In these latter regions, biotechnologists have not been able to convince consumers that recombinant crops are safe, both for humans and for the environment. As a result, the EU and the U.K. have enacted legislation that requires labeling for foods containing recombinant crops. Nervousness among food processors and retailers over consumer distrust in biotechnology has resulted in the disappearance of foods containing recombinant crops from grocery shelves in the U.K. Although some consumer surveys have shown a slight lessening of public antipathy to food biotechnology, it is unlikely that recombinant foods will reappear in Europe in the near future.

The Scope of Food Biotechnology

15

Will it harm the environment? Is is safe to eat? New technology is risky!

How does it benefit me? Consumer concerns

Consumer rejection of foods containing transgenic plants

FIGURE 1.5 Factors leading to consumer hostility toward foods containing transgenic plants.

The situation is very different in the U.S., Canada, and Argentina, where large acreages of recombinant soybeans, corn, and canola are sown. These countries do not have labeling legislation, although they do permit, with some restrictions, voluntary labeling of foods containing recombinant crops. Antibiotechnology activist groups have had limited success in raising public opposition to food biotechnology, but numerous surveys have indicated that there is a substantial number of North Americans that are strongly opposed to the presence of transgenic plants in food, particularly if the food does not carry labels indicating the presence of GMOs. Most polls have also indicated a strong desire to introduce mandatory labeling of GMOcontaining foods. So far, there has not been a ground swell of consumer activism to place pressure on the governments of the U.S. or Canada to introduce mandatory labeling, but that could change in the near future. Thus, the biotechnology industry in North America cannot afford to be complacent with respect to consumer activism and GMOs. The main areas of concern for consumers and antibiotechnology activist groups are the potential risks for human and environmental safety (Figure 1.5). In terms of human safety, a common perception is that GMO-containing foods have been inadequately tested for the presence of unpredicted allergens or toxins. This has been exacerbated by the fact that many of the companies involved in commercialization of transgenic plants have not released to the public the data used to establish that their transgenic crops are safe for human consumption. Many groups also feel that recombinant DNA technology is intrinsically dangerous, because it constitutes “unnatural” mixing of genes that would never occur using normal reproductive processes. The biotechnology industry’s attempts to reassure the public about human safety issues have largely been unsuccessful, partly because of continued insistence by the industry that human safety can be assessed using the process of substantial equivalence. This process involves a thorough comparison of levels of nutrients, toxins, and vitamins in a transgenic plant and the plant from which it was derived.

16

Introduction to Food Biotechnology

If these compounds are present in similar levels, the transgenic plant is considered to be substantially equivalent to the plant from which it was derived. The concept of substantial equivalence has been attacked by antibiotechnology activists and by a small number of influential scientists (e.g., the committee of the Royal Society of Canada that recently released an assessment of safety risks associated with food biotechnology). Scientists agree, though, that we lack alternative models for assessment of the safety of transgenic plants and that substantial equivalence has been successful to date. Transgenic plants have not caused human illness despite widespread consumption of foods containing transgenic plant products. Another point of contention among activist groups and some scientists (particularly ecologists) is that the environmental safety of transgenic crops has been insufficiently assessed. For example, there has been much discussion in lay and scientific circles about the possibility of herbicide-resistant crops becoming troublesome weeds, or of passing their novel genes to weed species. This is often referred to as the “superweed” problem. Insect-resistant crops have come under much criticism, particularly after studies were published in 1999 and 2000 showing that monarch butterflies were harmed by the ingestion of large amounts of corn pollen from transgenic corn expressing an insecticidal protein. Most government regulatory agencies (e.g., the U.S. Environmental Protection Agency) contend that such risks are carefully assessed during the regulatory process. However, the question of environmental risks continues to be a potent source of consumer hostility to GMOs, particularly in Europe and the U.K. The “What’s in it for me?” question is also an important component of consumer hostility. Many people in the biotechnology industry hope that the introduction of transgenic crops targeted to increase consumer health (e.g., functional foods with enhanced levels of health-promoting compounds) will lead to greater public acceptance of GMOs. As an analogy, consider the microwave oven. It is unlikely that microwave ovens would have been embraced so quickly by consumers if they had not offered significant benefits.

VII. FOOD SECURITY Food nourishes the body, and the production, processing, and distribution of food is crucial to global food security. It is also a crucial part of every nation’s economy and political stability. Industrialized nations have an abundant supply of high-quality and diverse food throughout the year. As a consequence of this richness in food, questions of food security typically focus on three issues: (1) ensuring that food is safe and is not contaminated with pathogens or pesticides; (2) public education to encourage sound nutritional practices; and (3) alleviation of poverty. However, in the developing world, food security is often low, because large numbers of people experience dire poverty. Food may be available, but that is irrelevant to those who lack the resources to buy it. This is a major problem in many countries that export food to richer nations. This illustrates the global inequity in food production and distribution that has been difficult to solve or alleviate, despite intense efforts in the latter half of the 20th century.

The Scope of Food Biotechnology

17

Biotechnology has the potential to increase food security. For example, many transgenic crops have decreased need of costly pesticides, because the crops themselves have been given the ability to fight off insects. Similar DNA-based technologies have also improved the range, sensitivity, and efficiency of diagnostic methods used to detect food-borne pathogens. An unfortunately small number of biotechnologists are developing low-cost diagnostic technologies that could be useful in developing countries. Biotechnology can also help fight poverty and malnutrition. β-carotene-enriched golden rice is the best example of this. Vitamin A deficiencies are widespread in the developing world and cause a staggering array of public health problems; golden rice has the potential to alleviate this problem. Biotechnology could also decrease the reliance of producers on chemical fertilizers, while retaining the benefits of “western” agriculture — high yields with reduced labor inputs. Although these beneficial results of biotechnology are real and substantial, the introduction of recombinant crops is a contentious issue throughout the developing world. Some countries (e.g., Sri Lanka) plan outright bans against the importation or planting of recombinant crops, whereas others (e.g., India) are attempting to develop their own biotechnology industry and are relatively receptive to recombinant crops. The key question is whether the use of recombinant crops will accentuate the positive or negative aspects of the “green revolution,” which led to enormous increases in agricultural productivity, but at the cost of increased economic disparity among farmers and increased reliance on technology and chemicals supplied by corporations from industrialized nations. Because most recombinant crops have been developed by corporate interests that are relatively uninterested in creating crops that are specifically tailored to agricultural problems in the tropics, where most developing nations exist, biotechnology may have a relatively small impact on this part of the world. However, increased western support of agricultural research in the tropics could lead to the development of transgenic crops targeted to specific agricultural problems in the developing world. This idea is supported by the observation that traditional plant breeding has sometimes benefited poor farmers. For example, a recent initiative to introduce drought-resistant maize seeds to Zambian communities was successful in decreasing the effects of drought on village farmers. It is possible that creative minds, particularly those in the developing world, will find biotechnological solutions for food production problems that will have a similar positive impact.

RECOMMENDED READING 1. Dunwell, J. M., Transgenic crops: the next generation, or an example of 2020 vision, Ann. Bot., 84, 269, 1999. 2. Daniell, H., Genetically modified food crops: current concerns and solutions for next generation crops, Biotechnol. Genet. Eng. Rev., 17, 327, 2000. 3. Henry, R. J., Using biotechnology to add value to cereals, in Cereal Biotechnology, Morris, P. C. and Bryce, J. H., Eds., Woodhead Publishing, Cambridge, 2000, chap. 5.

18

Introduction to Food Biotechnology 4. Etherton, T. D. and Kris-Etherton, P. M., Recombinant bovine and porcine somatotropin: safety and benefits of these biotechnologies, J. Am. Diet. Assoc., 93, 177, 1993. 5. Roller, S. and Dea, I. C. M., Biotechnology in the production and modification of biopolymers for foods, Crit. Rev. Biotechnol., 12, 261, 1992. 6. Adams, M. R. and Moss, M. O., Food Microbiology, 2nd ed., Royal Society of Chemistry, Cambridge, 2000. 7. Glazer, A. N. and Nikaido, H., Microbial Biotechnology: Fundamentals of Applied Microbiology, W. H. Freeman, New York, 1995. 8. Hoover, D., Chassy, B. M., Hall, R. L., Klee, H. J., Luchansky, J. B., Miller, H. I., Munro, I., Weiss, R., Hefle, S. L., and Qualset, C. O., IFT expert report on biotechnology and foods: human food safety evaluation of rDNA biotechnology derived foods, Food Technol., 54, 53, 2000. 9. Käferstein, F. K., Motarjemi, Y., Moy, G. G., and Quevado, F., Food safety: a worldwide public issue, in International Food Safety Handbook: Science, International Regulation, and Control, van der Heijden, K., Younes, M., Fishbein, L., and Miller, S., Marcel Dekker, New York, 1999, chap. 1. 10. Klijn, N., Weerkamp, A. H., and de Vos, W. M., Application of molecular detection and identification techniques in the study of the ecology of food associated microorganisms, in Progress in Microbial Ecology, Martins, M. E., Ed., Brazilian Society for Microbiology, Sao Paulo, Brazil, 1997, 214. 11. Wildman, R. E. C., Nutraceuticals: a brief review of historical and teleological aspects, in Handbook of Nutraceuticals and Functional Foods, Wildman, R. E. C., Ed., CRC Press, Boca Raton, FL, 2001, chap. 1. 12. Wenzel, G., The future role of bio-technology and genetic engineering, in Food Security and Nutrition: The Global Challenge, Kracht, U. and Schulz, M., Eds., St. Martin’s Press, New York, 1999, chap. 22. 13. Swaminathan, M. S., Toward a food-secure world, in Food Security: New Solutions for the Twenty-First Century, El Obeid, A. E., Johnson, S. R., Jensen, H. H., and Smith, L. C., Eds., Iowa State University Press, Ames, 1999, chap. 6. 14. Domoney, C., Mullineaux, P., and Casey, R., Nutrition and genetically engineered foods, in Nutritional Aspects of Food Processing and Ingredients, Henry, C. J. K. and Heppel, N. J., Eds., Aspen Publishers, Gaithersburg, MD, 1998, chap. 6. 15. Ye, X., Al-Babili, S., Klöti, A., Zhang, J., Lucca, P., Beyer, P., and Potrykus, I., Engineering the provitamin A (beta-carotene) biosynthetic pathway into (carotenoidfree) rice endosperm, Science, 287, 303, 2000. 16. Bertini, C., Food security: international dimensions, in Food Safety, Sufficiency, and Security, Special publication no. 21, Council for Agricultural Science and Technology, Ames, Iowa, 1998, 38.

2

Tools of the Trade

I. THE HEART OF BIOTECHNOLOGY: CELL BIOLOGY It is not easy being a biotechnologist. Biotechnologists need to be highly conversant in cell and molecular biology and comfortable with advanced topics specific to microbial, plant, or animal physiology. Many aspects of food biotechnology (e.g., functional food research) also require a sound understanding of food chemistry, as well as human nutritional epidemiology and physiology. As if that was not enough, food biotechnologists need to understand how market forces, regulatory bodies, and international trade can influence development of new technologies. It is beyond the scope of this book to supply all the necessary nutritional, food science, and biological background to become a proficient food biotechnologist. The emphasis in this chapter is on laying a solid foundation in microbiology and cell biology that will be helpful throughout the book. The reason for focusing on these topics is that microbes are tremendously useful to biotechnologists — they help us move bits of DNA from cell to cell, they produce many valuable biochemicals, and they are able to transform the ordinary into the extraordinary (e.g., grapes into wine). However, the main groups of microbes (viruses, bacteria, and fungi) have individual uses in biotechnology that are related to fundamental aspects of their structure and behavior. Hence, we will examine each group of microbes. Because DNA manipulation is so common in modern biotechnology, it is also important to understand the basics of how DNA works. We will also discuss some of the techniques that are used to manipulate DNA.

II. BACTERIA A. BACTERIAL GROWTH Bacteria are central to many aspects of food biotechnology (Figure 2.1). Bacterial enzymes, amino acids, vitamins, and polysaccharides are directly used in food processing, and they are also key participants in gene cloning and other molecular biological procedures. Bacteria are also the most frequent cause of food-borne illness, so development of efficient diagnostic tools for the detection of pathogenic bacteria is one of the major goals of food biotechnology. Thus, biotechnologists need to be aware of both the positive and negative aspects of bacteria. Why do bacteria have this dual (good/bad) nature? One major reason (Table 2.1) is that bacteria tend to have simple growth requirements; many need only carbon (e.g., glucose), mineral nutrients (e.g., nitrogen), and water. Consequently, they grow very well in most foods. The presence of bacteria in food commodities, soil, air, and water sources makes it difficult to eliminate them from food. For this reason, 19

20

Introduction to Food Biotechnology Enzymes

Amino Acids

Organic Acids (e.g. lactic acid)

BACTERIA

Vitamins

Food Additives (e.g. Xanthan gum)

FIGURE 2.1 Useful products of bacterial metabolism.

TABLE 2.1 Characteristics of Bacteria that Are Relevant to Food Safety and Food Biotechnology Characteristic

Relevance to Biotechnology

Relevance to Food Safety

Ease of growth Rapid growth

Cheap to grow in large scale Cheap and easy to grow in large scale or small scale Many opportunities for production of useful compounds Opportunity for the development of diagnostic tools to detect toxins Some microbes can be used in food preparation and processing Useful for genetic manipulations

Can grow in most foods Can quickly grow in many foods

Diverse physiology Toxin production Safe use in food Genetic structure

Difficult to assess safety Important cause of food-borne illness Difficult to assess for novel microbial foods Dangerous traits can move among bacteria

food microbiologists and engineers have developed many strategies for inhibiting microbial growth in food, including the application of organic acids, salt, heat treatments, freezing, and refrigeration. The ability to grow in simple media is also a blessing in certain contexts. Many bacteria can be easily grown at a laboratory scale, using nutrient broths and semisolid substrates such as agar or in large bioreactors (large-volume vessels specialized for the growth of microbes). Biotechnologists frequently use cheap “feeds” (materials used as a source of nutrients for the large-scale growth of microbes) such as molasses to supply all the nutrient requirements to bacteria grown in bioreactors. Thus, it is relatively easy to grow bacteria at a large scale. This partly explains the wide use of bacteria to produce products such as enzymes for the food industry. Many bacteria are capable of rapid growth. This characteristic is mainly due to their small size and simple cellular structure. Bacteria lack organelles and generally make fewer compounds than eukaryotic cells; thus they can produce required structures and compounds more quickly than a typical eukaryotic cell. Bacteria such as

Tools of the Trade

21

Escherichia coli can divide by binary fission in 30 min or less. This allows a single cell to produce a visible colony containing billions of cells in less than a day. Binary fission is simpler than mitosis, partly because of the presence of a single chromosome. Eukaryotic organisms, with their multiple chromosomes, require the complex process of mitosis to ensure that the cells resulting from division are genetically identical. In bacteria, this is a relatively simple process, allowing rapid growth. Again, this is both good and bad news for the food industry. It means that bacterial contamination of food can result in quick growth, resulting in large populations. This is a serious matter, particularly if the bacterium has pathogenic tendencies or produces toxins. On the other hand, the rapid growth rates of bacteria are useful to microbial biotechnologists interested in obtaining useful products from bacterial cultures. Rapid growth makes it easier to study bacteria. Consider a tube with a suspension of bacteria (Figure 2.2). This suspension can be diluted and spread over an agar (semisolid) medium in a petri dish. If the dilution is correct (i.e., not too dilute and not too concentrated), then each cell will be well separated from other cells. If these cells are then allowed to grow, each will produce a visible colony. Within each colony, the cells are genetically identical “clones.” This ability to quickly grow visible colonies from single cells is essential to many gene cloning procedures (see Chapter 3).

B. PHYSIOLOGICAL DIVERSITY Another useful characteristic of bacteria is their great physiological diversity. Virtually all natural organic compounds can be used as a carbon source by one bacterium or another, although each species tends to specialize (e.g., Erwinia carotovora can degrade many plant cell components). This diversity is useful to biotechnologists because the enzymes that bacteria use to degrade organic compounds often can be used to transform foods. The most important example for the food industry is the transformation of plant-derived starch into an array of compounds with a wide variety of uses as food ingredients (Table 2.2). The enzymes that drive these transformations are obtained from bacteria or fungi. Other products of bacterial metabolism, such as polysaccharides, vitamins, amino acids, organic acids, and lipids, are also useful. For example, the dairy industry depends heavily on the ability of lactic acid bacteria to efficiently ferment carbohydrates, releasing lactic acid as an end product. These bacteria are aerotolerant anaerobes (they can tolerate oxygen, but do not use it). They are completely dependent on fermentation to produce ATP (adenosine triphosphate) for cell growth and maintenance. As a consequence, they have become highly efficient at fermentation, and some strains, particularly those of Lactococcus lactis, are able to quickly produce enough lactic acid to drive the pH of milk from 6 to 4.5 within hours. The metabolic diversity of bacteria is also exploited during cheese ripening and sauerkraut production. In both of these processes, there is a succession of bacterial growth; one bacterium (e.g., Leuconostoc in sauerkraut) will first decrease the pH of the substrate (shredded cabbage for sauerkraut production). This creates conditions favorable for the growth of other bacteria, which then flourish for a period of

22

Introduction to Food Biotechnology A

B

C

FIGURE 2.2 Growing bacterial clones (genetically uniform colonies) from a suspension of cells. A suspension of cells (A) is spread over an agar plate containing the appropriate nutrients (B). Each cell grows into a colony of genetically uniform bacteria (C).

TABLE 2.2 Commercial Uses of Starch and Glucose-Derived Products Product

Use

Key Enzyme

Source of Enzyme

Microbe

Glucose Fructose Maltodextrin

Sweetener Sweetener Thickener

Glucoamylase Glucose isomerase α-Amylase

Aspergillus niger Streptomyces spp. Bacillus licheniformis

Fungus Bacterium Bacterium

Tools of the Trade

23

time. With sauerkraut, Lactobacillus plantarum spp. grow, and this is a crucial part of sauerkraut production, because this bacterium produces compounds that contribute to the flavor of the final product. Many other traditional fermented foods have a similar dependence on numerous species of bacteria (and fungi). For a bacterium to be exploited, it must be found and characterized. Traditionally, this has been achieved through screening processes, whereby a large number of bacterial isolates are tested for the presence of a desired attribute (e.g., production of a specific enzyme). Once a useful bacterium is isolated, it is identified and characterized further. For example, its ability to grow under a range of environmental conditions is assessed. This screening process is tedious, time consuming, and expensive. However, a number of developments in the last 20 years have made the process of microbial discovery much more efficient. Genomics — the study of entire genomes of organisms — is progressing at a rapid rate, fueled by rapid improvements in DNA sequencing technology. The DNA of at least 30 bacterial species has been fully sequenced. As more and more bacterial genomes are sequenced and studied, it will become easier to identify bacteria that have biotechnological potential. Bioinformatics becomes important in this context. This word refers to the use of DNA or protein sequence data to understand structure–function questions. For example, if one needs to find an organism that produces a heat-stable enzyme that breaks down pectin, and if enough sequence information is available, one can simply use computer software and databases to find candidate organisms that are thermophilic and produce pectinases. It is essential to have an orderly understanding of bacterial taxonomy to fully exploit their biotechnological potential. Current thinking divides all organisms into three domains: Archaea, Eubacteria, and Eukarya, which diverged early. All eukaryotic organisms are classified as Eukarya; Archaea and Eubacteria contain prokaryotic organisms that are fundamentally distinct from each other (e.g., cell wall structure is different in Archaea and Eubacteria). Most of the bacteria that are useful to biotechnologists are Eubacteria, but it is likely that Archaea will be used more in the future. Many Archaea are extremophiles that can tolerate extreme conditions of heat or salinity. Their enzymes often have similar extreme tolerances, which may prove to be useful in many biotechnological processes.

C. BACTERIAL GENETICS The basic mechanisms of flow of genetic information (i.e., DNA replication, transcription, and translation) are similar in bacteria and eukaryotes but are less complex in bacteria. For example, eukaryotes have a set of chromosomes (one set for haploid cells and two sets for diploid cells); each chromosome is made up of a doublestranded DNA molecule as well as numerous proteins (e.g., histones). In contrast, bacteria have a single chromosome, with few associated proteins. Also, mechanisms of gene regulation are generally simpler and more fully understood than in eukaryotes. For these and other reasons, gene expression is more easily modified in bacteria. Another advantage is that plasmids (small, circular DNA molecules that are replicated by the host cell’s DNA replication machinery) are common in bacteria but

24

Introduction to Food Biotechnology

Bacterium A

Plasmid Chromosome

Pilus

B

C

D

FIGURE 2.3 Bacterial conjugation. A conjugative plasmid in a host cell (A) contains genes that drive formation of a pilus (B) that joins the cell to another cell. A copy of the plasmid (C) is transferred to the other cell via the pilus (D).

rare in most eukaryotic cells. Plasmids are important to bacteria because they often contain useful genes (e.g., for antibiotic resistance), and certain plasmids, referred to as conjugative plasmids, contain genes that allow transfer of the plasmid from one cell to another through the process of conjugation (Figure 2.3). Initially, molecular biologists thought that conjugative plasmids would be useful tools for producing recombinant bacteria, but the problem with this approach is that recombinant genes may “jump” to other bacteria if the recombinant bacterium is released into the environment. For this reason, conjugating plasmids are not used for gene cloning. However, plasmids that are incapable of conjugation are useful DNA vectors in gene cloning. DNA from other sources can be easily inserted into plasmids, and the resulting recombinant plasmids can then be introduced into host bacteria such as E. coli (see Figure 1.1). The rapidity of bacterial cell division then allows the quick production of large amounts of plasmids, which can easily be purified from bacteria such as E. coli. This is possible because plasmid vectors contain an ori sequence (usually about 300 bases long) that is recognized by proteins that initiate DNA replication. Plasmids that lack an ori will not be replicated, resulting in loss of the plasmid from the bacterial population, because the plasmid will not be passed on to dividing cells.

Tools of the Trade

25

This is one of the reasons for the rarity of plasmids in eukaryotic cells — they are not usually recognized by the eukaryotic replication machinery. In bacteria, plasmids replicate independently of the chromosome; the frequency of plasmid replication varies widely. Thus, some plasmids are present in high numbers (high-copy-number plasmids) in each cell, whereas others are present in low numbers (low-copynumber plasmids). Generally, the former are preferred for cloning procedures. Because bacterial plasmids are recognized by bacterial polymerase, the cell must expend energy and materials toward plasmid replication. Thus, plasmids exert a metabolic load on host cells. Cells that have useless plasmids put their host cells at a disadvantage, compared to cells that lack plasmids. This leads to the loss of plasmids from bacterial populations, unless there is a gene on the plasmid that gives the host cell a competitive advantage over other bacteria. This is why plasmid vectors usually have antibiotic-resistant genes. If one wants to select cells that have a plasmid, exposing the cells to the antibiotic will result in death or growth suppression of cells that lack the plasmid. This strategy is central to most gene-cloning techniques (see Chapter 3, Section IV).

III. FUNGI ARE ALSO USEFUL AND VARIED A. GENERAL CHARACTERISTICS

OF

FUNGI

“Fungus” describes a large range of physiologically and structurally diverse microorganisms. Fungi are eukaryotic saprotrophs; they grow by breaking down organic compounds and utilizing the products of this decomposition as a source of carbon and energy. Fungi are directly important to the food industry (Table 2.3) as foods (e.g., mushrooms) and through the ability of fungi to transform food commodities. Economically, the most important industrial use of fungi is in the production of alcoholic beverages. Saccharomyces cereviseae is usually used for this purpose, although other yeasts, filamentous fungi, and bacteria share its ability to rapidly ferment carbohydrates to form ethanol. S. cereviseae is used to produce beer, wine, and virtually all distilled beverages. Its natural habitat is on the surface of sucroserich fruits (e.g., grapes), and it is well adapted for the fermentation of sucrose into ethanol and carbon dioxide. Yeasts are unicellular fungi, and S. cereviseae is the best understood and most widely used yeast. Because of its extensive history as a model cell for research on eukaryotic cell biology and genetics, S. cereviseae is probably the best-understood eukaryotic organism. This, combined with its long history of safe food use, has led to numerous applications in the food industry. Its most useful characteristics are rapid growth under both anaerobic and aerobic conditions, efficient production of carbon dioxide and ethanol, and “clean” metabolism (i.e., it can be used in food and beverage production without adding strong off flavors or odors). Various species of Aspergillus and Penicillium are extensively used in the food industry, because of their ability to transform basic commodities such as soybeans into value-added products such as soy sauce, and because they can secrete proteins in large amounts. These are filamentous fungi; they form extensive networks of filaments (hyphae) that have a width of one cell (Figure 2.4). Filamentous fungi are

26

Introduction to Food Biotechnology

TABLE 2.3 Importance of Fungi to the Food Industry Fungus

Characteristic

Relevant Food

Fruit body formation Ability to grow on cheese

Food mushroom Blue and camembert cheese

Recombinant yeasts with chymosin gene Aspartic proteases Lactase production Ethanol production CO2 production Rapid growth Enzyme production Citric acid production Flavor production

Production of chymosin (rennet) for cheese Cheese production Alleviation of lactose intolerance Fermented beverages Bread making Direct use as food Starch processing, juice processing Common food ingredient Soy sauce

Negative Effects Aspergillus flavus and others Penicillium digitatum and others

Mycotoxin production Invasion of plant tissues

Botrytis cinerea and others

Invasion of plant tissues

Contamination of various foods Cause of spoilage of fruits, grains, and vegetables Cause of plant disease in the field

Positive Effects Agaricus bisporus and others Penicillium roquefortii and Penicillium camembertii Kluyveromyces lactis and Aspergillus nidulans Mucor miehei and others Kluyveromyces lactis Saccharomyces cereviseae Saccharomyces cereviseae Saccharomyces cereviseae Aspergillus niger Aspergillus niger Aspergillus oryzae

efficient degraders of organic matter, because they can secrete hydrolytic enzymes and their hyphal colonies can quickly colonize organic matter. Amylases and proteases are two examples of useful enzymes that are produced by aspergilli. Unfortunately, not all fungi are so benign or useful. Many fungi produce toxic metabolites (mycotoxins). Aflatoxins, produced by Aspergillus flavus and Aspergillus parasiticus, are the most worrisome of the mycotoxins, because of their potency as mutagens and carcinogens, even at low concentrations. A. flavus is a common contaminant on peanuts and cereal grains. Consequently, aflatoxin contamination of food and animal feeds is an important food safety issue. Mycotoxins are particularly troublesome in the developing world, because: (1) fungi grow particularly well under warm, moist conditions, which are common in the tropics; (2) storage technology (e.g., refrigeration) is less widespread than in industrialized nations; and (3) developing nations frequently lack the economic resources to use diagnostic technology to detect fungal and mycotoxin contamination. Fungi are also a cause of great economic losses throughout the world, through the spoilage of stored food and the production of plant disease in the field. Historically, fungi have caused widespread epidemics of plant disease. For example, in the 1950s, North American wheat farmers faced large yield losses from wheat stem rust; this epidemic was finally stopped through the breeding of resistant wheat varieties, a process that continues to be important today. Fungal pathogens that attack field and harvested crops are currently an economic burden to both farmers and food

Tools of the Trade

27

FIGURE 2.4 Formation of a fungal colony from a single spore. The spore germinates to form a one-cell-wide filament (hypha). The hypha elongates and forms branches, leading to an interconnected colony of hyphae. Clusters of intertwined hyphae are mycelia.

processors; fungicide application is frequently necessary to control diseases both pre- and post-harvest. How can biotechnology help control these pathogens? Numerous efforts are underway to develop transgenic crops that have increased resistance to field diseases. This technology also has the potential to help protect plants from post-harvest diseases as well. Biocontrol, the use of organisms to prevent pathogens from attacking food plants, is enjoying increasing success and popularity partly because consumers throughout the world are pressing for decreased use of pesticides on food crops. Biotechnology is also useful in the detection and identification of fungal plant pathogens and mycotoxins. Antibody-based technologies have been particularly successful in this context — a number of antibody-based tests for mycotoxins in foods are commercially available. The development of diagnostic tests requires a thorough understanding of fungal taxonomy. The classification of fungi is complex, confusing, and currently in flux, but several basic divisions are well accepted. We now know that several of the groups traditionally included in the fungi are only distantly related to other fungi. This has led to a distinction between “pseudofungi,” which are more closely related to certain protist groups, and “true fungi,” which includes most of the economically important fungi. Both pseudofungi and true fungi can be troublesome plant pathogens. For example, the oomycete pseudofungus Phytopthera infestans, the biological cause of the Irish potato famine in 1850, is one of the most destructive plant pathogens. It is most closely related to heterokont protists, which includes diatoms and brown algae. In contrast, true fungi are phylogenetically more closely related to plants and animals than to pseudofungi.

28

Introduction to Food Biotechnology

The true fungi are divided into three divisions: Zygomycota, Ascomycota, and Basidiomycota. Most molds belong to Zygomycota or Ascomycota, whereas most mushrooms are members of Basidiomycota. Zygomycetes are quite different from other fungi. Their hyphae lack cross walls; this is one factor that allows them to grow rapidly over moist substrates. Zygomycetes are important agents of spoilage of fruits and other foods with high water activity. This group also contains useful fungi, particularly in the genera Mucor and Rhizopus. For example, Rhizopus oligosporus is used to transform soybeans into tempeh, a popular food in southeast Asia. Ascomycetes include most of the fungi important to food biotechnology (e.g., S. cereviseae), as well as a small number of prized food mushrooms (e.g., truffles and morels). Mycologists have also created an artificial group (Deuteromycota or Fungi Imperfecti) to contain isolates that lack a sexual stage in their life cycle. It has always been suspected that most of these species belong to either Ascomycota or Basidiomycota, and molecular techniques are increasingly used to determine the proper placement of species. Unfortunately, the current taxonomy of these fungi is confusing. Many species have two names: (1) the anamorph, the name classified under Deuteromycota, and (2) the teleomorph, the name that is given to the fungus if it can be classified under Ascomycota or Basidiomycota. This situation is relevant to food biotechnology because most of the fungi used in biotechnology belong to Deuteromycota. Thus, some species of Penicillium (anamorph) are also classified within the genus Eupenicillium (teleomorph). In general, though, anamorph names are usually used in reference to important fungi such as Penicillium, to minimize confusion.

B. THE USE

OF

FUNGI

IN

RECOMBINANT DNA TECHNOLOGY

Yeast and filamentous fungi are widely used in gene cloning, especially when the main aim of cloning the gene is to produce large amounts of recombinant proteins. Fungi are eukaryotes and share many of the organelles of plant and animal cells. These groups also share many intracellular processes. For example, messenger RNA (mRNA) transcripts in eukaryotes are often edited before translation (Figure 2.5), because eukaryotic genes typically contain noncoding regions. These introns are sequences of DNA within genes that are not translated into protein. Introns must be removed from mRNA transcripts before the mRNA leaves the nucleus and is translated in the cytoplasm. The remaining DNA sequences (exons) must also be spliced together to make an intact mRNA transcript before leaving the nucleus. This editing (removal of introns and splicing of exons) does not occur in eubacteria. This makes it difficult to clone eukaryotic genes in bacteria such as E. coli, although this difficulty can be alleviated through the use of cDNA (“copy” DNA) libraries (see Chapter 3, Section IV). Fungi can edit mRNA transcripts, but the editing process is slightly different than in mammals and plants. For this reason, cDNA libraries are frequently the method of choice, even when cloning a plant or animal gene into a yeast or a filamentous fungus. Another fundamental difference between eukaryotic cells and eubacteria lies in the process of protein secretion. Protein secretion in eubacteria is relatively simple; a hydrophic signal sequence allows the protein to pass through the plasma membrane.

Tools of the Trade

29

DNA

mRNA EXON

INTRON

EXON

INTRON

EXON

Translation

FIGURE 2.5 Gene splicing in eukaryotes. A typical eukaryotic gene contains coding and noncoding regions. The mRNA transcript includes exons (coding regions) and introns (noncoding regions). Introns are cut out of the transcript, and the exons are then spliced together.

The signal sequence is then enzymatically removed. In contrast, proteins destined for secretion by eukaryotic cells are produced in ribosomes associated with the rough endoplasmic reticulum (RER). After translation, the proteins enter the RER and undergo processing, during which carbohydrate chains (glycosylation) are added to the polypeptide (Figure 2.6). This processing, which takes place in the RER and the Golgi apparatus, is usually essential for the function of the secreted protein. Secreted proteins travel from the RER to the Golgi apparatus within membrane-bound vesicles (transport vesicles) and are released from the Golgi in similar vesicles (secretory vesicles) that fuse with the cell membrane, releasing their contents outside of the cell. Fungi are often able to process secreted proteins in a similar way as mammalian cells. Therefore, they are frequently a better choice for the expression (transcription + translation) of mammalian proteins than bacteria. Such production of recombinant mammalian proteins can be useful in the context of food production and processing; chymosin (rennet) and bovine growth hormone are two examples of this (see Chapter 3, Sections XI and XII). Some mammalian proteins are not processed properly in fungal cells, sometimes because of incorrect folding of the polypeptide after synthesis (in eukaryotes, specific proteins are involved in helping proteins achieve their proper, folded conformation). In such cases, the best alternative is to produce the protein in mammalian cell culture. In comparison to bacteria, the main disadvantage of using fungi is that they are physiologically and genetically more complex. Getting DNA into fungi is also more difficult, and fewer plasmid vectors are available for the transformation of fungal cells. Despite these shortcomings, fungi are increasingly popular, both as tools of basic research and as key players in biotechnological development.

30

Introduction to Food Biotechnology mRNA

Transcription of gene in the nucleus

Translation by ribosome associated with endoplasmic reticulum

ER lumen Ribosome

Protein passes into the endoplasmic reticulum

N-linked glycosylation

Transfer to Golgi apparatus

Golgi lumen

O-linked glycosylation

Secretion from cell

Secretory vesicle Plasma membrane

FIGURE 2.6 Processing of secreted proteins in eukaryotes. mRNA transcripts from the nucleus are translated by ribosomes associated with the rough endoplasmic reticulum (RER). The polypeptide then passes through the RER membrane into the lumen. Carbohydrate chains are added to NH2 groups of asparagine residues (N-linked glycosylation). The polypeptide is then transported via a membrane-bound vesicle to the Golgi apparatus, where carbohydrate chains are added to hydroxyl groups of certain amino acids (e.g., serine). This is referred to as O-linked glycosylation. Finally, the polypeptide (a glycoprotein) is secreted from the cell. The protein is enclosed within a membrane-bound vesicle that is pinched off from the Golgi. This secretory vesicle then fuses with the plasma membrane, releasing the protein into the external environment.

IV. VIRUSES: USEFUL PARASITES A. THE NATURE

OF

VIRUSES

Viruses are usually considered to be harmful parasites of cells, because of the enormous loss of life and economic productivity that they cause. Viruses are also important pathogens of crop plants and food animals, and the recent foot-and-mouth epidemic in Britain illustrates the ability of viruses to throw entire food production systems into chaos. As with bacterial and fungal pathogens, one of the benefits of biotechnology has been improved diagnostic tests for the detection and identification

Tools of the Trade

31

Envelope Spike Capsomere Nucleic acid

A

B

FIGURE 2.7 Structure of viruses. In nonenveloped viruses (A), the nucleic acid is surrounded by a capsid composed of protein subunits (capsomeres). In enveloped viruses (B), the capsid is enclosed by a membrane. The membrane usually has embedded protein spikes.

of viral pathogens. This is particularly relevant to food-borne viruses; despite widespread agreement that viruses are frequent causes of food-borne illness, most of the relevant viruses cannot easily be identified in food samples. Diagnostic biotechnology is expected to address this deficiency in the future. Viruses are also useful. All organisms are hosts to viruses, and many viruses are adept at moving pieces of DNA from one host cell to another. Therefore, it is not surprising that viruses have been used in cloning projects, where the prime object is to move specific segments of DNA from one organism to another. Viruses lack cellular structure and rely on host cells for reproduction. They can replicate only if they are able to infect the proper cell type and use the cell’s machinery (enzymes and metabolites such as nucleotides and amino acids) to produce new virus particles. Viral particles (Figure 2.7) are made up of a protein capsid enclosing the viral nucleic acid, which can be RNA or DNA. The nucleic acid contains all the genetic information required for successful infection of a host cell. Viral infection usually follows the following pattern (Figure 2.8): 1. The virus attaches to a host cell. This is usually a highly specific interaction. The virus will attach to a host cell, but not usually to a nonhost cell. The virus usually binds to specific proteins (receptors) in the host cell membrane. 2. The virus enters the cell. With regard to bacterial viruses (bacteriophage), only the viral nucleic acid enters the cell. In viral infections of eukaryotic cells, the entire viral particle enters, often through endocytosis. After entry of an entire particle, the particle is “uncoated.” This process results in breakdown of the capsid, releasing the viral nucleic acid into the cytoplasm. 3. The virus takes over the cell. Normal cellular metabolism is disrupted, and cellular enzymes, ribosomes, and metabolites are used to synthesize new viral proteins and nucleic acids. These components are then assembled to make new viral particles.

32

Introduction to Food Biotechnology

A

B

C

D

E

F

FIGURE 2.8 Infection cycle of a virus in a eukaryotic cell. The virus attaches to a host cell (A), enters the cell (B), is uncoated (C), directs the cell to make new viral components (D), assembles new particles (E), and causes lysis of the cell (F), thus releasing the new virus particles.

Tools of the Trade

33

4. New virus particles are released from the cell. Bacteriophage are released through lysis of the cell, whereas viruses of eukaryotes are sometimes released through lysis and sometimes through budding, which may not be lethal to the host cell. Enveloped viruses acquire their envelope through budding — the viral particle is enclosed in part of the plasma membrane that is pinched off. This is broadly similar to the process involved in secretion of proteins through the formation of secretory vesicles (see Figure 2.6). This infection cycle is often called the lytic cycle. The lysogenic cycle is similar, but includes a protracted stage when the virus is inactive and the viral nucleic acid is incorporated into the DNA of the host cell. One of the common points of confusion in biotechnology centers on the use of bacteriophage lambda (λ) in gene cloning. In introductory microbiology courses, this virus is usually taught as an example of a virus that has a lysogenic cycle. However, when λ is used in cloning procedures, genes required for lysogeny are usually removed. These strains quickly lyse susceptible cells after infection. See Chapter 3, Section VIII for an explanation of the use of viruses in gene cloning.

V. DNA: THE HEART OF BIOTECHNOLOGY A. DNA STRUCTURE Deoxyribonucleic acid (DNA) and its sister compound ribonucleic acid (RNA) are vital components of many biotechnological applications. The molecular biology revolution that has occurred in the last 20 years and created so many new biotechnological opportunities is fundamentally based on the ability to precisely manipulate DNA. Therefore, it is essential for biotechnologists to have a thorough understanding of DNA. The prime role of DNA is to act as a reservoir of genetic information. This is possible because of the following structural features of DNA (Figure 2.9). • DNA is a double helix made up of two antiparallel strands. • Each strand is made up of a backbone of deoxyribose monosaccharides linked covalently through phosphate bridges. • Each deoxyribose unit is linked covalently to a base consisting of either adenine (A), guanine (G), cytosine (C), or thymine (T). • Two antiparallel strands, through hydrogen bonding between adjacent base pairs, can form a stable double helix. • Hydrogen bonds form between complementary base pairs (C–G and A–T). • Three linear bases on a strand code for a specific amino acid — this allows a linear sequence of bases on a strand of DNA to code for a linear sequence of amino acids on a polypeptide. Each group of three bases is a codon. The antiparallel (upside down relative to each other) nature of the two strands is important in allowing a stable configuration of the double helix. The orientation of each strand is best described using 3′ and 5′ notation (see Figure 2.9). Close

34

Introduction to Food Biotechnology

5’

3’ B

D

B

B

D

B

D

B

D

B

P D P

P D

B

P

P D

B

P

P D

B

3’

P

P

5 1 B 4 D 2 3

D

P

5’

FIGURE 2.9 The structure of DNA. The backbone is made up of phosphate (P) and deoxyribose (D). The bases (B) interact through hydrogen bonds.

examination of the structure of the two strands shows that each phosphate molecule is attached to carbon 3(3′) of one deoxyribose and to carbon 5(5′) of the adjacent deoxyribose. At one end of a strand, there is a deoxyribose with a 3′ that is not attached through a phosphate molecule to another deoxyribose. At the opposite end of the strand, there is a deoxyribose with a 5′ end that is free. Thus, each strand has a 5′ end and a 3′ end. As we will see, the orientation of the strands is important to the function of DNA. The DNA double helix can be reversibly separated into two single strands. This happens naturally during DNA replication and transcription and is driven by a specific enzyme. Heating also results in separation of the strands (melting), as does increasing the pH of the solution (e.g., through addition of NaOH). The latter process is referred to as denaturation. Both methods are used extensively in molecular biology. If either condition is relaxed (i.e., by cooling or by removing the NaOH), double-stranded DNA will re-form, through the association of complementary strands. This re-annealing of single-stranded DNA is called hybridization when the two strands are from different sources. Hybridization often involves of the use of a single-stranded DNA probe that anneals to a target strand of DNA. Hybridization can occur between two strands of DNA, two strands of RNA, or one each of DNA and RNA, as long as they are complementary. Hybridization of primers (short segments of single-stranded DNA) to longer strands of DNA is also an essential part of the polymerase chain reaction (PCR).

Tools of the Trade

A

B

35 3’

5’

5’

3’

3’ 5’ *

t C

*

FIGURE 2.10 Replication of DNA. The double-stranded DNA (A) separates into single strands. A primer (*) anneals to one of the strands (B), and DNA polymerase (filled circle) extends the new strand (C), starting with the primer, and adding bases complementary to the template strand (t). The other original strand would also act as a template.

B. DNA REPLICATION When a cell divides, it must ensure that both daughter cells have identical genetic structure; in other words, each cell must have the same sequence of DNA. This is achieved through DNA replication. In both bacteria and eukaryotes, this is accomplished through a suite of proteins and enzymes that separate the DNA into single strands, allowing DNA polymerase to use each strand as a template to build a new complementary strand (Figure 2.10). The following characteristics of DNA replication are particularly important; they are crucial to various techniques of molecular biology, including PCR: • DNA polymerase cannot start to build a new strand of DNA unless it has an RNA primer that is complementary to a sequence on the template strand. DNA polymerase is able to extend a complementary DNA strand from this primer and later replaces the RNA primer with DNA. DNA primers are often used to control the starting point of DNA replication. • The new DNA strand is always synthesized from the 5′ to the 3′ end. As noted in the discussion of plasmids, DNA replication in cells also requires a specific DNA sequence that acts as an “origin of replication” (ori) on the chromosome (eukaryotic cells often have multiple sites of origin on each chromosome).

36

Introduction to Food Biotechnology

A

P

O

S

T

FIGURE 2.11 Structure of a eubacterial gene. Upstream of the structural region (S), there is a promoter (P) and possibly an operator (O) or activator (A) sequence. Downstream of the structural region, there is a termination region (T).

C. TRANSCRIPTION

OF MRNA

The end product of transcription is a single-stranded mRNA molecule that is complementary to a sequence on one of the strands of DNA. This strand of DNA is the coding or sense strand, and the strand of DNA that is not transcribed is the noncoding or nonsense strand. The sense strand forms a template for the construction of a complementary strand of mRNA, in a similar fashion as strands of DNA act as templates for the construction of new DNA strands during DNA replication (see Figure 2.10). The driving force behind transcription is RNA polymerase. This enzyme is similar to DNA polymerase in that it requires an RNA primer before it can build a complementary strand and it builds the new strand from the 5′ to the 3′ end. The direction that RNA polymerase moves is downstream and the opposite direction is upstream. To understand transcription, it is useful to examine the structure of the coding strand of DNA (Figure 2.11). It consists of a sequence of DNA (the structural region) that codes for mRNA that will be translated into a protein. On either side of this region are regions of DNA that are not transcribed or translated into protein. The region directly upstream of the protein-coding region is essential for transcription. It is the promoter region. This sequence of bases is required for transcription; it allows RNA polymerase to recognize start points. The promoter in eubacteria is always upstream of the protein-coding region and consists of discrete DNA sequences that interact directly with the RNA polymerase protein. This is an example of protein–DNA interaction, which is a crucial element in many aspects of gene activity and regulation. In eubacteria, one of the subunits (e.g., the sigma factor) of RNA polymerase is often required to assist in binding of RNA polymerase to the promoter. This gives bacteria a large-scale regulatory tool. Sigma factor binds to the promoter of certain genes, whereas other factors bind to the promoters of other genes. Therefore, the particular DNA-binding factor that is present will greatly influence the pattern of gene activity. Directly downstream of the promoter, many eubacterial genes also have an operator region, which is important in regulation of transcription. Repressor proteins bind to the operator region (another example of DNA–protein interaction) and block the action of RNA polymerase. This negative regulation is relieved, in some cases, by the presence of an inducer molecule that binds to the repressor protein, preventing its binding to the operator.

Tools of the Trade

37

Eubacterial genes sometimes use positive regulation acting on activator-binding sites that are upstream of the promoter region. Activator proteins bind to this region and enhance binding of RNA polymerase to the promoter, thus increasing the frequency of transcription of the gene. The promoter region tells RNA polymerase where to start transcription, but other sequences are required to indicate the point when transcription should end. This is done by terminator sequences. In eukaryotes, transcription is more complex. Numerous additional proteins, called transcription factors, are required to help RNA polymerase bind to the sense strand of DNA. Promoters are present upstream of the protein-coding region and generally have the same purpose as in eubacteria. However, eukaryotes do not use negative regulation. Instead, they tend to rely on positive regulation, through the use of enhancers, DNA sequences that are usually upstream from the promoter. The promoter in eukaryotic genes often consists of a number of short DNA sequences (modules) interspersed with less important sequences. One module, the tata box, tells RNA polymerase where to start transcription. Many eukaryotes are multicellular organisms, with a complex hierarchical structure made up of cells, tissues, and organs. This hierarchy is often reflected by specificity of promoters; for example, many promoters in plant genes are tissue specific. This is important to biotechnologists; for example, when transferring a gene into a plant, it is often possible to restrict activity of the new gene to a specific tissue or organ, through the use of specific promoters.

D. EDITING

OF

RNA TRANSCRIPTS

IN

EUKARYOTES

As mentioned in Section III of this chapter, one of the fundamental differences between eubacteria and eukaryotes is that RNA transcripts of eukaryotes are edited to remove introns, before they can function as mRNA. This splicing process is usually (but not always!) absent in eubacteria and archaea. When it does occur in prokaryotes, it occurs through a fundamentally different mechanism than in eukaryotes. Eukaryotic RNA transcripts (pre-mRNA) are spliced by a complex of RNA and protein (ribonucleoprotein) called a spliceosome, whereas prokaryotic splicing is done autocatalytically (i.e., the RNA molecule itself has the ability to edit out introns and join the exons).

E. TRANSLATION

OF MRNA INTO

PROTEIN

Translation describes the process of protein synthesis by ribosomes, using a mRNA transcript. The combined processes of transcription and translation lead to the expression of proteins. How does translation proceed? In eubacteria, mRNA associates with ribosomes soon after transcription has begun. Recall that ribosomes are made up of several subunits of ribosomal RNA (rRNA) associated with specific proteins. In an orderly fashion, transfer RNA (tRNA) molecules charged with amino acids interact with the ribosome and the mRNA, resulting in a chain of amino acids whose sequence is determined by the base sequence of the mRNA. Examination of

38

Introduction to Food Biotechnology

3’

5’ Leader

AUG

Stop T

Protein

FIGURE 2.12 Structure of eubacterial mRNA. T = trailer.

a typical eubacterial mRNA molecule (Figure 2.12) reveals that the AUG start codon marks the beginning point of translation (i.e., the point where the ribosome begins to construct a polypeptide). To the left of this point (toward the 5′ end), there is a leader region that is not translated. It is essential for translation, though, because it contains the Shine–Dalgarno sequence, part of the ribosome-binding site, which is required for the initial binding and interaction of the mRNA with a ribosome. The ribosome-binding site also overlaps into the initial part of the protein-coding sequence of the mRNA. Near the 3′ end of the mRNA, a stop or nonsense (because it does not code for an amino acid) codon (e.g., UAA) signals to the ribosome the end of the polypeptide. The new polypeptide and the mRNA are then released from the ribosome. To the right of the stop codon (toward the 3′ end) there is a noncoding region referred to as the trailer. In eukaryotes, the mechanism of translation at the ribosomes is similar, but the structure of the mRNA is different. Enzymes in the nucleus add methyl (CH3) groups to specific points of several bases at the 5′ end of the leader region of the molecule; this is the methylated cap. This cap apparently helps the mRNA bind to ribosomes in the cytoplasm. Another processing event occurs in the nucleus — poly A polymerase adds 100 to 200 adenine nucleotides onto the 3′ end of the mRNA transcript. The function of this poly A tail is unclear, but may be related to transport of mRNA from the nucleus to the cytoplasm.

F. POSTTRANSLATIONAL PROCESSING

OF

POLYPEPTIDES

In eubacteria, polypeptides leave the ribosomes and achieve their final folded conformation, often with the help of small proteins called molecular chaperones. This is an essential process, because protein function depends on the three-dimensional conformation of the protein. Disulfide bond formation is a crucial part of this folding process and is partly dictated by the cytoplasmic environment — if the cytoplasm contains an abundance of compounds that have strong reducing power (i.e., a “reducing” environment), disulfide bond formation is less likely to occur. Disulfide oxidoreductases, oxidative enzymes that catalyze the formation of disulfide bonds, are also essential to this process in both bacteria and eukaryotes. Eubacteria have simple cellular structure. When a protein is produced in the cytoplasm, it will function in one of three locations: in the cytoplasm, in the cell membrane, or outside of the cell (secreted proteins). Proteins that are destined for the membrane have a specific signal sequence of amino acids at their aminoterminal end (this is the end that is first synthesized by the ribosome; the opposite end is the carboxy-terminal end). As a secreted protein passes through the membrane, a protease removes the signal sequence. A protein that contains a signal sequence is called a preprotein.

Tools of the Trade

39

In eukaryotes, proteins must be sorted and targeted to various locations in the cell (nucleus, endoplasmic reticulum [ER], vacuole, etc.). This complex process is also guided by amino-terminal (N-terminal) signal sequences. Biotechnologists are primarily interested in secreted proteins; these proteins are synthesized by ribosomes closely associated with the ER. Regions of ER with many associated ribosomes are known as rough ER. Secreted proteins have a signal sequence that leads to rapid interaction with signal recognition particles (SRPs). This occurs before the polypeptide has left the ribosome. The SRP binds to docking proteins in the membrane of the ER, forming a complex that directs the new polypeptide into the lumen of the ER. When the polypeptide is released from the ribosome, it begins its journey from the ER to the Golgi apparatus and out of the cell. While it is in the ER, the important process of glycosylation begins. This involves the addition of oligosaccharides to specific points in the polypeptide. These carbohydrate chains are added either to the free amino group of asparagine (N-glycosylation) or to the OH group of serine, threonine, or lysine (O-glycosylation). N-glycosylation begins in the ER and is completed in the Golgi, whereas O-glycosylation occurs only in the Golgi (see Figure 2.6). The exact pattern of glycosylation (i.e., the length of each chain and the nature of the subunits of the chain) varies among eukaryotes; for example, the pattern in yeasts is quite different from that in mammalian cells. Secreted proteins travel from the ER to the Golgi via transport vesicles that are budded off from the ER. From the Golgi apparatus, proteins are transported to the membrane via secretory vesicles. The membrane of a secretory vesicle fuses with the cell membrane, releasing its contents to the outside of the cell. For many proteins, though, this is not the end of posttranslational processing. Many proteins exist in a pro form (e.g., pro-insulin). Such proteins must be modified (i.e., some of the amino acid sequences must be removed) before the protein achieves its final, active configuration. In some cases (e.g., prochymosin), the amino acid sequence is able to catalyze this final processing step; in others, proteases are required. In some cases, environmental factors (e.g., strong acid in the stomach lumen) drive the final conversion to an active protein. Because of this final processing step, secreted proteins are associated with confusing jargon. For example, chymosin (rennet) is initially produced as preprochymosin. When the signal sequence has been removed, it is then called prochymosin, and when it has been cut into its final active form, it is called chymosin.

G. RELEVANCE

OF

DNA

TO

BIOTECHNOLOGY

The preceding sections may seem like overkill — why do food biotechnologists need to know so much about DNA? Many biotechnological processes involve the transfer of genes from one organism to another. It should now be clear that if this transfer occurs between organisms that are quite distantly related (e.g., from a mammalian cell to a eubacterial cell), problems are likely to occur. Will the promoters, termination signals, ribosome-binding sites, and signal sequences work in the new cell? Will the RNA transcript be correctly edited before translation? Will the new cell have the machinery for proper posttranslational processing? These are crucial questions, and one of the goals of Chapter 3 is to address them.

40

H. WORKING

Introduction to Food Biotechnology WITH

DNA

One of the major barriers confronting newcomers to biotechnology is understanding the jargon associated with the day-to-day manipulation of DNA. Four essential techniques will be described here. 1. Purification of Nucleic Acids It is often necessary to separate DNA from other cellular constituents. One important reason for this is that cells contain nucleases that cut nucleic acid polymers into small fragments or individual nucleotides. Fortunately, it is relatively easy to purify DNA from cells. The cell must first be lysed (broken). With animal cells, this is easily accomplished through the use of solutes such as sucrose (high levels of sucrose cause water flow out of cells, resulting in bursting of the cell), membrane solubilizing agents such as sodium dodecyl sulfate (SDS), and proteases. Cells with walls (e.g., fungal, plant, and bacterial cells) may require additional or different treatments. For example, yeast cells are usually treated with cell-wall-degrading enzymes to yield sphaeroblasts (cells bound by a membrane only), which lyse easily. Once cellular contents have been released, a chemical such as phenol is added; this causes denaturation of any proteins present. Organic solvents such as chloroform are added to solubilize lipids. This mixture is then centrifuged, which results in three distinct layers: one made up of phenol, one aqueous layer, and a layer of chloroform. The aqueous layer is carefully removed. The next step is to add cold ethanol; this precipitates DNA and RNA, which can then be pelleted through centrifugation and resuspended in a small amount of water or buffer. Such DNA samples are usually treated with RNAse (a nuclease that degrades RNA) to eliminate contaminating RNA. RNA samples must be handled very carefully, because RNA is less stable than DNA, and RNAses are much more common in the environment. Gloves can prevent contamination of RNA samples with RNAses present on skin. Using water (in buffers, etc.) that contains chemical inhibitors of RNAses is another preventive measure. 2. Gel Electrophoresis When working with DNA, it is often necessary to separate fragments of different lengths. Also, determining the length of DNA fragments is frequently useful. Both aims can be accomplished using gel electrophoresis. This technique involves the addition of a sample containing DNA to a gel (agarose or polyacrylamide) that is suspended in a buffer. DNA is loaded into small slots cut into the gel. An electrical current is then applied, and the charge that DNA carries (due largely to the phosphate groups of the backbone) will move the DNA fragments toward the positive pole (Figure 2.13). However, the fragments do not move through the pores of the gel at equal rates; small fragments flow rapidly, whereas large fragments are impeded by the gel structure, and move slowly. Thus, fragments of different sizes are separated. A number of samples can be added to each gel, creating discrete lanes. Fragments of known length can be added to certain lanes, allowing measurement of DNA fragment length in the other lanes. Once electrophoresis is completed, DNA in the

Tools of the Trade

41

Power supply

- pole (cathode) Buffer reservoir Slots

Gel

Buffer reservoir

+ pole (anode)

FIGURE 2.13 Gel electrophoresis. The gel is submerged in buffer, and DNA fragments migrate toward the anode (indicated by the arrow).

gel can be visualized by adding ethidium bromide to either the gel or the DNA sample. Ethidium bromide binds to DNA and fluoresces visibly when exposed to ultraviolet light (Figure 2.14). Molecular biologists measure DNA size in units of base pairs or kilobase (kb) pairs (thousands of base pairs). Gel electrophoresis can accurately separate and measure small fragments of DNA, but for segments larger than 40 kb, special techniques (e.g., pulsed-field gel electrophoresis) are required. 3. Blotting and Hybridization Blotting and hybridization techniques are commonly used to detect sequences of DNA. For example, southern blotting can determine if a specific DNA sequence is present in a sample of DNA. This is often done after gel electrophoresis. Blotting refers to the wicking of DNA from a gel or other substance onto a membranous filter. Water is encouraged to flow through the gel and the membrane filter by immersing the gel in a solution and placing a paper towel or similar blotting paper over the filter. As water flows to the blotting paper, DNA flows with it but is trapped on the membrane filter. Once on the filter, the DNA can be denatured into single strands through immersion in a solution of NaOH. Then, a probe can be added (after removing the NaOH). A probe consists of a specific sequence of DNA or RNA that is labeled in some way. Two common methods of labeling are to introduce radioactive isotopes to the probe or to covalently link a fluorescent compound to the probe.

42

Introduction to Food Biotechnology 1

2

3

4

5

6

FIGURE 2.14 Separation of DNA fragments by gel electrophoresis. Lane 1 contains fragments of known molecular weight. This ladder can be used to measure the size of DNA fragments in lanes 2 through 6.

If a sequence in the DNA sample is complementary to the probe, hybridization will occur (Figure 2.15). The unhybridized probe is then washed from the filter, and hybridized probes can be detected using either radiation-sensitive film for radioactive probes or fluorescence (production of visible light upon exposure to ultraviolet light) for fluorescent probes. The above procedure can also be used to detect specific mRNA transcripts. This is referred to as northern hybridization. RNA or DNA probes can be used in both northern and southern hybridizations. Hybridization can also be accomplished without gel electrophoresis. Dot blots, for example, are made by placing drops of a sample containing DNA onto a nitrocellulose filter. Labeled probes can then be added. Unhybridized probes are washed away, and hybridization can then be detected. Dot blots offer a quick test for the presence of a particular sequence of DNA and are widely used. 4. DNA Sequencing Since the advent of the human genome project, DNA sequencing has become part of everyday language. However, it is useful to gain a deeper understanding of the process, because of its central importance to understanding and using genes. If one knows the sequence of a gene, it can be compared to genes from other organisms, and its amino acid structure can be deduced through our understanding of the genetic code. This often leads to substantial insights into protein structure and function and may be highly relevant to biotechnological applications of the gene. Sequencing also has directly practical applications, in that it allows the design of gene probes

Tools of the Trade

43 Target DNA GACGATCTCGAT + Probe CTGCTAGAGCTA

* *

*

GACGATCTCGAT CTGCTAGAGCTA

* *

*

FIGURE 2.15 Hybridization of a DNA probe to a DNA sample. The DNA sample must first be converted to single strands. The probe then is able to hybridize with the complementary sequence in the sample. The probe is labeled by covalently linking it to an enzyme of fluorescent compound (*).

that can be used to detect the gene, as well as the design of primers to allow amplification of the gene using PCR. DNA sequencing can occur through either the Maxam–Gilbert or the Sanger dideoxy method. We will describe the Sanger dideoxy method here because it is the most commonly used. • Obtain a sample of DNA fragments containing the desired gene. • Divide the sample into four tubes. • To each tube, add nucleotides of the four bases plus a synthetic analogue of one of the bases. For example, the A tube would contain guanidine triphosphate (GTP), cytosine triphosphate (CTP), thymidine triphosphate (TTP), ATP, and dideoxy ATP. All of the nucleotides are labeled to allow detection of either radioactivity or fluorescence (see blotting section, above). • Add DNA polymerase and DNA primers to each tube. DNA polymerase uses the triphosphate nucleotides ATP, CTP, TTP, and GTP to add nucleotides to the primer, creating a new strand of DNA. However, it is unable to add dideoxy ATP, CTP, TTP, or GTP onto a strand. Thus, when DNA polymerase is forced to use a dideoxy nucleotide, strand elongation stops. • Consider the A tube; it contains ATP as well as dideoxy ATP. As DNA polymerase adds nucleotides to the new strand, it will add ATP whenever it “sees” thymine in the template DNA (i.e., the DNA fragment that is being sequenced). If it adds a normal ATP, then the strand will continue

44

Introduction to Food Biotechnology

to elongate. However, if it tries to add a dideoxy ATP, then strand elongation will cease. Consequently, wherever T (thymine) occurs in the template, dideoxy ATP will stop strand elongation some of the time. This will result in a series of fragments in the A tube, each resulting from termination of strand elongation by dideoxy ATP. • Separate the fragments in each tube by high-resolution gel electrophoresis, using a polyacrylamide gel. • You can then detect the fragments, either by the use of autoradiography (radiation-sensitive films) or through fluorescence (many automated gene sequencers use fluorescence). • “Read” the DNA sequence, starting from either the top or the bottom of the gel.

RECOMMENDED READING 1. Madigan, M. T., Martinko, J. M., and Parker, J., Brock Biology of Microorganisms, 8th ed., Prentice Hall, New York, 1997. 2. Sahm, H., Prokaryotes in industrial production, in Biology of the Prokaryotes, Lengeler, J. W., Drews, G., and Schlegel, H. G., Eds., Verlag, Stuttgart, 1999. 3. Deacon, J. W., Modern Mycology, 3rd ed., Blackwell Science, Oxford, 1997. 4. Lodish, H., Berk, A., Zipursky, S. L., Matsudaira, P., Baltimore, D., and Darnell, J. Molecular Cell Biology, 4th ed., Freeman, New York, 2000. 5. Lewin, B., Genes, 7th ed., Oxford University Press, Oxford, 2000. 6. Wolff, R. and Gemmill, R., Purifying and analysing genomic DNA, in Analyzing DNA: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 1997, chap. 1.

3

Gene Cloning and Production of Recombinant Proteins I. WHAT IS A RECOMBINANT PROTEIN?

Consumers generally recognize the importance of protein to their diet. However, few people are aware of the important role of specific proteins as food modifiers or additives. Amylases, for example, are used to transform starch into sweeteners such as glucose. Other amylases are used by the brewing industry to help convert starch into fermentable carbohydrates. Proteases can be used to tenderize meat and to coagulate milk during cheese making. Usually, such enzymes are derived from plant, animal, or, more commonly, microbial sources. These are “natural” sources of the proteins. In contrast, recombinant proteins are derived from organisms that do not naturally make the protein. The ability to make the protein is transferred into the organism. In molecular terms, the DNA that codes for the protein is transferred. As an example, consider a protease that is naturally found in pineapples. Using recombinant DNA technology, we can isolate the DNA coding for this enzyme from the rest of the pineapple DNA and introduce it into a bacterium such as Escherichia coli. Then, we can grow large amounts of the bacterium and purify the enzyme from the culture. The protease would then be a recombinant protein. In the parlance of molecular biology, it would also be called a heterologous protein, because it is produced by an organism different from the original source organism. Food biotechnologists usually produce recombinant proteins in bacteria or fungi, because of the ease of large-scale culture of these organisms. In contrast, mammalian or human cell lines are often the source of recombinant proteins for human therapy; in the future, therapeutic proteins will also be produced in whole plant and animal systems. A more common food-related application of recombinant proteins is in the development of transgenic (genetically modified) crops. The recombinant protein improves the crop, benefiting food producers, processors, or the consumer. This important and controversial application of recombinant DNA technology will be discussed in Chapter 4.

II. WHY BOTHER MAKING RECOMBINANT PROTEINS? To make a recombinant protein, one has to transfer the specific gene from one organism to another. This is not always easy. Why, then, do people bother? Why not just obtain the protein from the organism that naturally produces it? 45

46

Introduction to Food Biotechnology

Sometimes it is not practical to use the natural producer. For example, human growth hormone is used to treat dwarfism, a condition that arises when children are congenitally deficient in human growth hormone. This specific protein is found only in humans; consequently, the only natural source of this hormone is from human cadavers. Unfortunately, only minute amounts can be obtained from cadavers. However, once the gene for human growth hormone was transferred to mouse cell lines, it could be produced relatively cheaply in large amounts. Bovine growth hormone (BGH), also known as bovine somatotropin (BST), is found in cows. Similar to human growth hormone, BGH cannot easily be isolated from its natural source. However, in the 1980s, the gene for BGH was transferred from cows to E. coli, and the resultant recombinant bacteria were able to produce intact BGH (rBGH, or recombinant BGH, to differentiate it from natural BGH). This made it cost effective to sell rBGH to dairy farmers. When rBGH is injected into cows, milk production usually increases. We will discuss this controversial recombinant protein further in Section XII of this chapter. Chymosin is another example of a protein that is available from either natural or recombinant sources. This enzyme has been used for thousands of years to increase the rate and extent of coagulation of milk during cheese production. Most of the cheese produced in North America is made using a recombinant form of chymosin. One advantage of recombinant chymosin is that it has more consistent effects on its substrate (casein in milk) than chymosin from rennet, the traditional source. Recombinant chymosin will also be discussed at the end of this chapter (Section XI), because the story of its development demonstrates many of the technical problems associated with making recombinant proteins in microbes. In some cases an enzyme may be readily available from plant or microbial sources, but the enzyme may have undesirable qualities. For example, several microbial proteases can be used instead of chymosin, but they tend to produce off flavors in cheese, because of excessive proteolytic activity. The enzymatic activity could theoretically be directly modified by changing the amino acid sequence of the active site. This can be done through site-directed mutagenesis, and the resulting protein would be considered a recombinant protein, even if it was produced by the organism that produced the original protein. The aim of this chapter is to explain the processes that are used to clone genes into bacteria and fungi and that allow the production of recombinant proteins from these cloned genes. These processes are important because of the increasing presence of recombinant proteins in the food supply and because many of the techniques described in this chapter are used in other kinds of food biotechnology. Understanding the basic methods of gene cloning is essential to many of the subsequent chapters of this book. Understanding the techniques of producing recombinant proteins is also crucial for a complete understanding of the controversies associated with recombinant food additives (e.g., recombinant chymosin) and recombinant plants (e.g., herbicide-resistant soybeans). One of the most controversial aspects of genetically modified foods is the inclusion of antibiotic resistant genes. Why are these genes necessary? A close look at gene cloning principles is necessary to answer this

Gene Cloning and Production of Recombinant Proteins

A

47

DNA from organism with the desired gene

B

Cut the DNA into fragments

C

Insert fragments into vector

D

Transform cells with vector

E

Use probe to identify cell with desired gene

F

Culture cells with desired gene

*

FIGURE 3.1 Shotgun cloning. In this example, the vector is a plasmid.

question. Gene cloning is the necessary first step toward production of a recombinant protein in a microbe or plant. The transfer of a gene from one organism to another can be conceptually divided into two processes: (1) cloning the gene and (2) transferring the cloned gene into the desired organism.

III. HOW AND WHY ARE GENES CLONED? (THE BIG PICTURE) To transfer a desired gene from one organism (the source) to another (the target), one must separate the gene from all other genes in the source organism. Otherwise, one loses the principal advantage of a recombinant organism — precise transfer of genes, rather than wholesale mixing of large numbers of genes. One way of achieving this goal is to use the following approach, commonly referred to as shotgun cloning (Figure 3.1):

48

Introduction to Food Biotechnology

1. Obtain DNA from the source organism (i.e., the organism that has the desired gene). 2. Cut the DNA using restriction enzymes. These enzymes cut DNA at specific sequences and allow the conversion of a large chromosome, or a number of chromosomes, into a set of linear segments of DNA. 3. Incorporate the DNA fragments into a vector. A vector is an agent that can replicate in target cells. Usually plasmids or bacteriophage are used as vectors when the target cell is a bacterium. For now, assume that the vector is a plasmid, a circular DNA molecule that is able to replicate in host cells (see Chapter 2, Section II). At this stage, we cannot differentiate between wanted and unwanted segments. Each segment will be incorporated into a plasmid. A small proportion of plasmids will have the desired gene, and the rest of the plasmids will have undesired genes from the source organism. 4. Move the vector into the target cells. Usually, at this stage, the target cell is E. coli. Plasmids can be moved into E. coli by exposing the cells to a treatment that opens pores in the cell membrane, allowing plasmids to diffuse into the cells. This process is usually inefficient, and it is necessary to eliminate cells that do not have plasmids. Plasmid vectors typically have an antibiotic-resistant gene, which acts as a marker or indicator gene. If the antibiotic is then added to the bacteria, only those bacteria that contain the plasmid will survive. The next step allows separation of cells with plasmids containing undesired genes from cells with plasmids containing desired genes. 5. Use a probe to isolate cells with plasmids that contain the desired gene; the most common is a DNA probe. This is a short sequence of DNA that is complementary to the desired gene. Because it is complementary, it will anneal to the desired gene, but it will not anneal to any other fragment of DNA. The probe must also be linked to a compound that can be easily detected; radioactive or fluorescent compounds are most popular. 6. Grow large amounts of cells containing the desired gene. These cells have the required genetic information for producing the desired recombinant protein. The protein can therefore be harvested from cultures of these cells. The desired gene has been successfully cloned. 7. The target cell should be able to efficiently synthesize large amounts of the protein coded for by the desired gene. Often, though, the target cell, for a variety of reasons, will not synthesize enough of the desired protein. You may need to repeat the procedure, transferring the desired gene to a cell that is better suited to the synthesis of large amounts of the protein. This procedure is referred to as subcloning and is conceptually similar to shotgun cloning. However, it is usually much easier than shotgun cloning, because the interfering “background” of the source organism’s genome is absent.

Gene Cloning and Production of Recombinant Proteins

A

49

Eco RI G A A T T C C T T A A G

A A T T C G G C T T A A

B

Rsa I G T A C C A T G

G T C A

A C T G

FIGURE 3.2 Effects of the restriction enzymes EcoRI (A) and RsaI (B) on DNA.

IV. GENE CLONING: THE DETAILED PICTURE A. RESTRICTION ENZYMES Restriction enzymes are classified as endonucleases because they cut DNA at points within the molecule. The crucial advantage they have over other endonucleases is that they do not cut DNA at random points. Instead, they cut at specific DNA sequences. For example, EcoRI will cut DNA whenever the enzyme encounters the sequence GAATTC (Figure 3.2). Therefore, the catalytic action of the enzyme is restricted to certain sequences of DNA. The following points about restriction enzymes are important to their use in gene cloning:

50

Introduction to Food Biotechnology

A DNA

B

C

FIGURE 3.3 Relationships between the length of a restriction sequence and the frequency of cleavage of DNA. The same DNA strand is depicted in each part of the figure. Each arrow indicates a point where the DNA strand is cut. RsaI (A) recognizes the sequence GTAC; EcoRI (B) recognizes GAATTC; and SfiI (C) recognizes GGCCNNNNNGGCC (N = any base).

• Each enzyme usually recognizes only one sequence. • Many restriction enzymes have been characterized. Some recognize short sequences (e.g., 4 base pairs), whereas others recognize long sequences. A strain of Streptomyces fimbriatus recognizes a 13 base pair sequence, an unusually long restriction sequence. The length of the recognized sequence is extremely important because that affects the frequency of cuts. For example, consider three subsamples taken from a sample of DNA (Figure 3.3). Each subsample is exposed to one of the following three restriction enzymes: RsaI, EcoRI, and SfiI. What happens? Most likely, RsaI, which recognizes a 4 base pair sequence will make a large number of cuts. EcoRI will make an intermediate number of cuts, and SfiI will make few cuts. Evidently, the biotechnologist’s choice of restriction enzymes is important; appropriate enzymes must be chosen whenever restriction enzymes are used in a cloning procedure. For each type of vector, there is a limit to the amount of DNA that can be inserted. The correct restriction enzyme will result in the correct fragment size for a particular vector. • Many restriction enzymes create “sticky ends.” These are “loose” single strands of DNA protruding at the cleavage site (see Figure 3.2A). Sticky ends are useful because complementary sticky ends will anneal to each other. This allows the joining of DNA from different sources. If each of two samples of DNA has been treated with the same restriction enzyme, and if this enzyme generates sticky ends, then, when the samples are mixed, annealing will occur between DNA strands from the two different samples. However, sticky ends from the original sample may also reanneal, because they are also complementary. Also, certain restriction enzymes create blunt ends (see Figure 3.2B) rather than sticky ends.

Gene Cloning and Production of Recombinant Proteins

51

• Usually, the DNA sequence recognized by a restriction enzyme is a palindrome. The sequence of DNA is identical in both strands, when read from the same orientation (e.g., from 5′ to 3′; see Figure 3.2). Restriction enzymes are isolated from bacteria. At first encounter, these enzymes appear to be unusual. Why would a bacterium make such unusual enzymes? It turns out that restriction enzymes are part of a bacterial defense system against bacteriophage. This system works because restriction enzymes degrade foreign DNA, such as phage DNA that has entered the cell. Of course, bacterial DNA always contains sites recognized by its own restriction enzymes. Specific enzymes recognize restriction sites and methylate them. This protects bacterial DNA from digestion by its own restriction enzymes.

B. PLASMID VECTORS Plasmids are vital tools for biotechnologists. Over the past 20 years, these circular strands of DNA have been extensively used to clone genes. Plasmids are useful because they are small, easily manipulated, and can be “grown” in large quantities. Their use in cloning is best explained with the following scenario: you have isolated a fungus from soil that produces an enzyme (amylase X) that is able to efficiently break down starch into valuable food ingredients. Unfortunately, the fungus is difficult to grow in large-scale culture, so you decide to try to move the gene (AmylX) coding for this enzyme into E. coli (Figure 3.4). You use pBR322, a commonly used plasmid vector. Examination of pBR322 (Figure 3.5) reveals that it contains an ori sequence, necessary for replication in E. coli, and two antibiotic-resistant genes. These antibiotic-resistant genes are vital in the cloning process, which will now be explained in detail. Step 1: Purify the DNA containing the desired gene. This is usually a straightforward procedure (see Chapter 2, Section V). One of the advantages of cloning procedures is that only small amounts of DNA are required. Therefore, if the organism that contains the desired gene is in short supply, or grows very slowly, this is not usually a debilitating problem. We will refer to this DNA as foreign DNA. Step 2: Purify plasmid DNA. This is done by growing E. coli containing pBR322, and then breaking open the bacterial cells to release cellular contents. Plasmid DNA can then be easily purified and separated from chromosomal DNA, because the small size of plasmids gives them physicochemical characteristics different from chromosomal DNA. Step 3: Cut both plasmid and foreign DNA with the same restriction enzyme. In this case, BamHI is a good choice. This enzyme cuts DNA at the following sequence: GGATCC. There are three reasons for this step: a. The fungus contains several different chromosomes, which are all present in the purified DNA sample. However, only one chromosome contains AmylX, and you want to clone only the portion of the chromosome that contains AmylX. If the DNA is cut into small pieces, each

52

Introduction to Food Biotechnology

Obtain DNA that contains the desired gene

Obtain plasmid DNA (pBR322)

Cut both ‘desired’ DNA and plasmid DNA with the same restriction enzyme

Mix DNA fragments, and ligate

Transform E. coli with ligated DNA

Select bacteria with plasmids Screen for plasmid with desired gene

Screen for plasmid with recombinant plasmids

FIGURE 3.4 Cloning of a gene into E. coli using pBR322.

Eco RI Bam HI

tet

amp

r

r

Pst I

ori

FIGURE 3.5 Structure of pBR322. There are two genes for antibiotic resistance: tetr and ampr.

Gene Cloning and Production of Recombinant Proteins

53

piece can theoretically be integrated into separate plasmids. It is then relatively easy to separate the plasmid containing the desired gene from plasmids containing the “noise” (the rest of the fungal DNA). b. It is necessary to cut the fungal DNA into small pieces because it is impractical to attempt to incorporate more than 4 kb (four thousand bases) of DNA into pBR322. This restriction applies to many plasmid vectors, but other vectors, such as lambda phage, cosmids, and yeast artificial chromosomes, can accommodate larger segments of DNA. c. BamHI creates sticky ends. This will allow the segments of foreign DNA to anneal to plasmid DNA. Note that there is only one BamHI restriction site (G↓GATCC) in pBR322 (i.e., there is only one sequence in pBR322 that BamHI recognizes and cuts). Therefore, the only effects of digestion with BamHI will be to convert the circular plasmid to a linear plasmid and to generate sticky ends. If there was more than one restriction site, pBR322 would be cut into several pieces, and it would no longer be a useful vector because each piece would likely be missing an ori, or an antibiotic-resistant gene. It is also important to note that the BamHI restriction site occurs within the tetracycline-resistant gene. This will become important in step 6. Step 4: Mix the DNA containing the desired gene with the vector DNA (Figure 3.6). Random collisions between pieces of DNA with complementary sticky ends will result in annealing, or joining, of DNA segments. This annealing is not permanent, unless DNA ligase is added, because simple annealing does not create covalent links between the sugar phosphate backbones of the segments. DNA ligase covalently welds (ligates) annealed segments of DNA. The main goal here is to join segments of the foreign DNA with plasmid DNA, forming recombinant plasmids. If this step is successful, you will obtain a population of recombinant plasmids, some of which have the desired gene and many of which have other segments of foreign DNA. In addition to recombinant plasmids, the population usually contains re-annealed plasmids that lack any inserted DNA. These re-annealed plasmids can be separated from recombinant plasmids in step 7. Step 5: Transform E. coli with the mixture of annealed and ligated DNA. The desired plasmids cannot be isolated unless they are introduced into a microorganism. The most common way to introduce plasmids and other DNA sequences into E. coli, and other bacteria, is via transformation. Transformation is a confusing word, because it has been given diverse meanings by different groups of biologists. In the context of gene cloning, transformation refers to the uptake of DNA from aqueous solution by microorganisms. Most microorganisms cannot do this; one notable exception is Streptococcus pneumoniae. This bacterium has specific proteins that allow DNA from other cells of S. pneumoniae to enter the cell. The DNA may then be integrated into the cell’s chromosome. E. coli, however, does not naturally take up DNA in this manner. Bilayer lipid membranes will not allow the diffusion of DNA into cells, and E. coli lacks the proteins that allow S. pneumoniae to take up DNA. Fortunately for biotechnologists,

54

Introduction to Food Biotechnology

G G A T C C

G G A T C C

C C T A G G

C C T A G G

GGATCC CCTAGG

A. BamHi cuts DNA and pBR322

G

G A T C C

C C T A G

G

B

GATC

C GATC

FIGURE 3.6 Annealing and ligation of pBR322 and a segment of DNA containing the desired gene. Both plasmid and DNA are cut with BamHI (A). They are then mixed together, allowing annealing of plasmid and DNA (B). Finally, ligase is added, to covalently bond the plasmid to the DNA insert (C).

E. coli can be treated (i.e., be transformed) to increase their ability to take up DNA. The most common transformation method for E. coli is to expose it to a dilute solution of cold CaCl2. This appears to transiently “freeze” lipid membranes, resulting in the formation of fissures or pores that are large enough to allow the diffusion of DNA into the cell. One disadvantage to this method is that it is extremely inefficient — only 1/106 of the bacterial cells will be transformed. An additional problem is that many bacterial species cannot be transformed using this method. Once the plasmid DNA is inside E. coli, it can be stored there indefinitely. DNA polymerase will bind to the ori region of the plasmid and will replicate it, preventing

Gene Cloning and Production of Recombinant Proteins

55

Master plate

Replica plate Identify missing colonies Subculture missing colonies

FIGURE 3.7 Separation of bacteria containing recombinant pBR322 from bacteria with nonrecombinant pBR322. Recombinant bacteria have a disrupted tetracycline gene and are sensitive to tetracycline, which is present in the replica plate but not the master plate. Once sensitive bacteria have been identified, they are subcultured from the master plate.

loss of the plasmid as the bacterium divides. The amount of replication that occurs varies widely among plasmids. Plasmids that replicate frequently are considered to be high-copy-number plasmids. PBR322 is a low-copy-number plasmid and is consequently not as popular as other, more frequently replicating plasmids, such as the pUC group of plasmids. Transformation will yield a mixture of bacteria. Most will remain untransformed and will lack plasmids. A small proportion of the transformed bacteria will contain plasmids with the desired foreign gene, and the remainder will contain plasmids with undesired sequences of foreign DNA. Bacteria that contain plasmids with inserted segments (inserts) of foreign DNA are considered to be recombinant. Step 6: Select bacteria transformed with plasmids. Untransformed bacteria can easily be removed by adding ampicillin. E. coli is susceptible to ampicillin, so the only bacteria that will survive in the presence of ampicillin are those that have been successfully transformed with pBR322. The bla gene, found in pBR322, codes for β-lactamase, which degrades ampicillin. Step 7: Screen for recombinant clones. Application of ampicillin results in a mixture of bacteria that have been successfully transformed. Even if steps are taken to prevent the re-annealing of plasmids without inserts, it is usually necessary to separate reannealed plasmids from plasmids containing inserts. Plate the mixture of transformed bacteria onto a solid agar medium containing ampicillin (the master plate; Figure 3.7). The bacteria must be diluted sufficiently, so that each colony grows from a single cell. Bacteria within each colony, then, are clones; they are genetically identical, because they share the same plasmid and the same insert within the plasmid. A sterile velvet cloth is then carefully pressed against the master plate and pressed onto an agar plate that contains tetracycline. This will result in the transfer of bacteria from each colony onto the tetracycline-containing plate. It is essential to maintain the orientation of the velvet cloth, so that the bacteria are pressed onto the

56

Introduction to Food Biotechnology

Add labeled probe Master plate

Replica on filter paper

Wash away unbound probe

Lyse colonies, releasing DNA

Detect bound probe

FIGURE 3.8 Screening of recombinant libraries using a DNA probe.

tetracycline-containing plate in the same pattern as the pattern of colonies on the ampicillin-containing plate. This plate is referred to as a replica plate. Bacteria that have the original, recircularized plasmid are able to grow on media containing tetracycline, because of the presence of the gene giving tetracycline resistance (see Figure 3.5). However, bacteria that have recombinant plasmids are killed by tetracycline, because they do not have an intact tetracycline-resistant gene. Insertion of foreign DNA at the BamHI restriction site (see Figure 3.5) breaks the tetracycline-resistant gene into two fragments, with the foreign DNA between them. This results in transcription of a gene that is not functional; i.e., it does not give the cell resistance to tetracycline. Therefore, cells with recombinant plasmids (recombinant clones) grow on the ampicillin-containing plate but not on the tetracyclinecontaining plate. Recombinant clones, once identified, are subcultured from the ampicillin-containing plate. The resulting mixture of recombinant clones is often referred to as a DNA library. Step 8: Screen the DNA library. Few of the recombinant plasmids in a shotgun DNA library contain the desired gene; most have inserts of undesired foreign genes. It is now necessary to screen the recombinant clones to identify those that contain the desired gene. The most common strategy is to use a DNA probe (Figure 3.8). This is a short segment of single-stranded DNA that is complementary to a portion of the desired gene. It must be labeled in some way. Often, radioactive elements such as 32P are incorporated into the probe, but nonradioactive alternatives are also widely used. Nonradioactive probes are either fluorescent or biotin labeled. Fluorescein and several proprietary compounds can be covalently linked to oligonucleotides; hybridization of these probes to DNA can be detected by applying ultraviolet light. Fluorescent compounds emit light of a longer wavelength (often in the visible spectrum) when exposed to ultraviolet light. Thus, hybridized probes will “light up.” Biotinlabeled probes work on a different principle; biotin is a vitamin that binds specifically to avidin (a protein found in eggs) or streptavidin (a bacterial protein). Avidin can be covalently linked to enzymes such as horseradish peroxidase that, when given synthetic analogues of their normal substrate, form a colored end product. Thus,

Gene Cloning and Production of Recombinant Proteins

57

hybridization can be detected through the formation of a complex consisting of DNA bound to biotin-labeled probe, which is in turn bound to enzyme-linked avidin. When the synthetic substrate is added, a colored end product will form. The first step in probing a library is to plate out the library on an agar medium (master plate). A filter paper is then pressed onto the master plate, removing bacteria from each colony while preserving the orientation of the master plate. Cells are then lysed by the application of alkaline chemicals (e.g., NaOH), which also denatures plasmid DNA, yielding single strands. This allows the single-stranded probe to anneal specifically to complementary single-stranded DNA on the filter paper. This is an example of hybridization (the annealed double-stranded DNA is a hybrid of one strand of probe DNA and one strand of plasmid DNA; see also Chapter 2, Section H). The probe is then added to the filter paper and, after incubation, the filter paper is washed to remove unbound (not annealed) probe. Hybridization is then detected via radioactive or fluorescent emission, depending on the type of probe. Once the desired recombinant colonies have been identified, the bacteria can be picked from the master plate and grown in large volumes. At this point, the gene has been cloned. Usually, biotechnologists will take further steps to confirm that the correct gene has been transferred to E. coli. Confirmation can be made by determining the DNA sequence of the insert or by testing for presence of the desired protein in cultures of the recombinant bacterium. DNA probes cannot be used if nothing is known of the DNA sequence of the desired gene. However, this is rarely the case; for most genes, sequence information is available for similar genes in other organisms. If such information is not available, then it may be necessary to use another method of library screening. If the desired gene is an enzyme, then it may be possible to screen clones for the desired enzyme activity.

V. THE cDNA ALTERNATIVE Shotgun cloning works well when the desired gene is from bacteria or other organisms that have a small genome. However, it is less successful when we try to clone genes from eukaryotes, especially eukaryotic plants and animals, which have particularly large and complex genomes. One reason for this lack of success lies in the large number of recombinant clones that must be screened (most screening processes are laborious and time consuming). However, there is a more fundamental problem. As discussed in Chapter 2, Section II.B, DNA sequences in eukaryotes are not directly transcribed and translated into protein. The primary mRNA transcript typically contains several segments (introns) that are “edited out” before translation. The sequences that survive editing (exons) are spliced together into a mRNA molecule that leaves the nucleus and is translated by ribosomes. This makes it difficult to use shotgun cloning for eukaryotic genes. If the entire DNA sequence for a eukaryotic gene is transferred to E. coli, the protein that is produced will be completely different from the native eukaryotic protein. E. coli lacks editing ability, and the entire primary mRNA transcript would be translated

58

Introduction to Food Biotechnology

Isolate mRNA from cells (tissues) that produce the desired protein Add Reverse Transcriptase DNA-RNA hybrids Digest mRNA and add Klenow fragment Double stranded DNA forms

Insert DNA sequences into a vector (e.g., a plasmid)

Screen the cDNA library for the desired clone

FIGURE 3.9 The use of reverse transcriptase to make a cDNA library. Reverse transcriptase uses mRNA as a template to produce a complementary strand of DNA. Other enzymes are then used to eliminate the mRNA and produce a double-stranded DNA molecule. The DNA segments can then be inserted into vectors, leading to construction of a cDNA library. This library can then be screened in the usual manner.

into protein. It is highly unlikely that such a protein would retain the properties of the original protein. Fortunately, there is a solution, which comes from an unlikely source: retroviruses. Retroviruses use RNA to store their genetic information. As part of their infection cycle, they convert the RNA to DNA, which is then incorporated into the DNA of the host cell. Reverse transcriptase catalyzes this conversion. This enzyme is used by biotechnologists to generate DNA from edited mRNA (Figure 3.9). The DNA produced is double-stranded and does not contain introns. This strategy is attractive because mRNA can usually be isolated from the organism that naturally produces the desired protein. Reverse transcriptase then uses the mRNA strands as templates to generate a complementary strand of DNA. This results in a hybrid molecule consisting of one strand of RNA and one strand of DNA. The mRNA can then be digested away using specific nucleases, and the Klenow fragment of DNA polymerase I will use the single-stranded DNA as a template for a complementary DNA strand. The Klenow fragment of DNA polymerase I is able to synthesize complementary strands of DNA from a DNA template, but it lacks some of the

Gene Cloning and Production of Recombinant Proteins

59

nuclease activity of intact DNA polymerase. It is used for a variety of molecular biology procedures. The combined action of reverse transcriptase and the Klenow fragment results in a double-stranded DNA molecule. The rest of the cloning procedure is similar to that described for shotgun cloning; the only difference is that the DNA used in step 3 is obtained from edited mRNA transcripts, not from the genomic DNA of the organism that naturally produces the desired gene. DNA produced from mRNA is referred to as cDNA (c = copy), and libraries obtained by this method are cDNA libraries. This approach has allowed the cloning of a large number of eukaryotic genes that could not be cloned using shotgun cloning.

VI. POLYMERASE CHAIN REACTION: A REVOLUTION IN CLONING A. OVERVIEW The last decade has seen an extraordinary increase in the use of polymerase chain reaction (PCR) for cloning and other molecular biological procedures. This revolution has occurred because PCR allows repeated replication (amplification) of specific sequences of DNA. One drawback is that information about the sequence of the desired gene is required. More specifically, the sequence at both ends of the gene must be known; these sequences are used to construct primers. The only other requirements are a heat-stable DNA polymerase, a mixture of nucleotides, and a thermal cycler, which is simply a device that can rapidly heat and cool a solution in a programmed sequence. The heat stability of the DNA polymerase is essential to the process. Fortunately, such enzymes can be isolated from a variety of prokaryotes that inhabit water close to geothermal vents; many of these organisms can grow in water at or above 100°C. PCR is often used to simplify gene cloning. If we know enough about a desired gene to construct primers, we can use PCR to locate the DNA coding for the gene and incorporate it into a plasmid vector. Cloning is completed by transformation of the recombinant plasmid into a suitable host. This works because PCR amplifies only the desired gene. Therefore, virtually all of the noise (i.e., the rest of the genome) is eliminated at an early stage of the process. DNA amplified from eukaryotic genomic DNA contains exons and introns, and consequently will not be useful. However, DNA coding only for exons can be obtained by starting with mRNA, using reverse transcriptase to convert the mRNA to cDNA. We can then use PCR to amplify the cDNA.

B. THE MECHANISM

OF

AMPLIFICATION

First, it is important to understand primers. Each primer is a short sequence (usually ~20 base pairs) of single-stranded DNA, and the base sequence of the primers is complementary to the flanking ends of the gene (Figure 3.10A). The other requirement is that the primers must be complementary to different strands. If the primers are identical to flanking ends of the same strand of DNA, the reaction will not work.

60

Introduction to Food Biotechnology A 3’ 5’

Desired gene CTAAGCT GATTCGA

Primer A: 5’ GATTCGA 3’

5’

GGCATCT CCGTAGA

Primer B:

3’

3’ 5’

GGCATCT

Heat to 90°C 3’ 5’

CTAAGCT

GGCATCT

GATTCGA

CCGTAGA

5’ 3’

Cool to 50°C

3’ 5’

5’

CTAAGCT GATTCGA

5’

GGCATCT 3’ 3’

GATTCGA

5’ GGCATCT CCGTAGA

3’

Heat to 72°C

3’ 5’

5’

CTAAGCT GATTCGA

GGCATCT CCGTAGA

GATTCGA

GGCATCT CCGTAGA

5’

5’ 3’

FIGURE 3.10 Amplification of DNA by polymerase chain reaction. In the first cycle (A), primers bind to each strand, allowing DNA polymerase to synthesize complementary strands. Note that each new strand extends past the desired region (shaded) in one direction. In the second cycle (B), primers bind again to each strand. However, note that one of the new strands runs only between the primers. The noise (undesired DNA) outside of the shaded region has not been amplified. In the third cycle (C), the amplified segments containing only the region between the primers are starting to dominate (numerically) segments containing regions outside the primers. This domination increases exponentially as the cycles continue.

Once we know the flanking sequences to a desired gene, we can simply construct (or buy) oligonucleotides of the correct sequence, which can then be used as primers. The synthesis of oligonucleotides (short segments of DNA or RNA with a defined base sequence) is an automated, relatively inexpensive procedure. Primers are added to the DNA sample that contains the desired gene, a heatstable DNA polymerase is added, and this reaction mixture is placed in a thermal

Gene Cloning and Production of Recombinant Proteins

61

B 3’ 5’

CTAAGCT GATTCGA

5’

GGCATCT CCGTAGA

Primer A: 5’ GATTCGA 3’

Primer B:

3’

GGCATCT

5’

Heat to 90°C 3’ 5’

CTAAGCT

GGCATCT

GATTCGA

CCGTAGA

5’ 3’

Cool to 50°C 3’ 5’

CTAAGCT GATTCGA

3’

GATTCGA

5’

5’

GGCATCT 3’

GGCATCT CCGTAGA

5’ 3’

Heat to 72°C

3’ 5’

CTAAGCT GATTCGA

GGCATCT CCGTAGA

GATTCGA

GGCATCT CCGTAGA

* 5’

5’

5’ 3’

FIGURE 3.10 (continued).

cycler. The PCR consists of a number of repeated cycles. The following steps are part of a typical PCR cycle: 1. The reaction mixture is heated to 90°C. This results in separation of double-stranded DNA into single strands. 2. The mixture is cooled to 50°C. This allows annealing of the primers to the single-stranded DNA templates. Note that the primers anneal to separate strands of DNA. 3. The mixture is heated to 72°C. This allows DNA polymerase to initiate DNA synthesis, starting at the primers. DNA polymerase requires a primer for initiation of synthesis of a new DNA strand, and it synthesizes the new strand in the 5′ to 3′ direction.

62

Introduction to Food Biotechnology

C 1st cycle

1st cycle products

2nd cycle products

3rd cycle products

FIGURE 3.10 (continued).

The cycle is then repeated 20 to 30 times. This results in amplification of the DNA sequence between the primers. Close examination reveals why other DNA sequences are not amplified. The orientation of the primers always directs DNA polymerase into the gene. After several cycles of PCR, the sequence between the primers is present in much higher numbers than the rest of the DNA sample. Eventually, after 30 to 40 cycles, the amplified DNA has been copied 109 times, effectively drowning out the original sample of DNA. Let’s go through the first few cycles, to illustrate how this works. In the first cycle (Figure 3.10A), the DNA is heated to 90˚C and the double-stranded DNA denatures into single strands. When the temperature decreases to 50°C, the primers

Gene Cloning and Production of Recombinant Proteins

63

anneal to complementary sequences. Note the orientation (5′ to 3′) of each DNA segment; remember that complementary sequences will anneal only if the two strands are antiparallel, and will not if they are parallel (same orientation). Note also that each of the template strands at this stage extends in both directions. What happens when we allow DNA polymerase to replicate new strands of DNA? DNA polymerase must have a primer; it requires a sequence of DNA that can be used as the initial segment of the new strand. By using designed primers, we control the starting point for DNA replication. Also, DNA polymerase can synthesize the new strand of DNA in only one direction. The new strand is synthesized in the 5′ to 3′ direction. So, after one round of replication, each of the original template strands has been amplified. The second cycle (Figure 3.10B) reveals the specific nature of the amplification. To keep things simple, consider what happens to only one of the newly synthesized strands. As in the first cycle the first step is to increase the temperature to 90°C, resulting in denaturation of the double-stranded DNA into single-stranded DNA. Then we decrease the temperature and allow annealing to occur. Note that only one of the primers has the correct orientation to anneal to this strand. DNA polymerase then synthesizes a new strand, starting from the primer. The new strand is (as always) synthesized in the 5′ to 3′ direction. The newly synthesized strand is a short segment that includes only the two flanking primer regions and the DNA sequence between the primers. The rest of the original segment of DNA is no longer present. This is the crucial feature of the PCR; it leads to selective amplification of the DNA between the primers. Subsequent cycles result in more and more of these short segments (Figure 3.10C), and eventually, after 20 or more cycles, they are present in large enough quantities that they can be visualized on an agarose gel. In other words, the DNA sequence between the primers has been amplified. After about 30 cycles, amplified DNA is present in about 1 million times as many copies as the original DNA.

C. PCR IS NOT PERFECT PCR has revolutionized molecular biology and can shorten and simplify many processes, such as gene cloning. However, problems often arise when using PCR. The most serious complication is DNA contamination. If extra DNA is present, the PCR will be unsuccessful if the contaminating DNA contains sequences complementary to the primers. Part of the contaminant DNA will be amplified, resulting in a mixture of desired and undesired amplified products. The DNA polymerase that is normally used in PCR is another source of problems, because it lacks proofreading ability. Most DNA polymerases can detect mismatched bases; for example, if the DNA template reads A, and DNA polymerase adds C to the new complementary strand, it recognizes the mistake, removes the incorrect base, and replaces it with the correct base. The enzyme usually used in PCR (TAQ™) is derived from the thermophilic eubacterium Thermus aquaticus. It lacks the normal proofreading ability of DNA polymerase. As a result, the wrong base is sometimes integrated into new strands. This happens in a random manner and is not usually a problem unless it occurs early in the cycle. Also, many applications of PCR aim to detect genes; in such applications, occasional mistakes in

64

Introduction to Food Biotechnology

DNA replication should not affect the amplification process nor reduce the ability of PCR to detect a gene. However, if the aim is to isolate a gene for the purpose of cloning it, even a rare mistake could eventually cause problems. For this reason, it is advisable to use a heat-stable DNA polymerase with proofreading ability when using PCR to obtain DNA for cloning. For example, the DNA polymerase of Thermococcus litoralis, a thermophilic Archaen, has proofreading ability. In the context of gene cloning, the biggest drawback of PCR is that it is much less effective when the sequence to be amplified is longer than 5 kb. This is a serious problem because many genes are longer than this. However, there are ways to address this, and PCR remains an extremely useful tool in gene cloning programs.

D. VARIATIONS

ON

PCR

Primers do not always bind to only one target. If another gene in the DNA sample has a similar sequence, it will also be amplified, resulting in a mixture of products. This problem can be sidestepped by amplifying the mixture of amplified DNA using a second set of primers. The second set of primers amplifies within the DNA amplified by the first primers. It is highly unlikely that a sequence that is amplified by both sets of primers exists within a sample. This approach is known as nested PCR. Hotstart PCR is another alternative method that is frequently used. It can be effective in solving mispriming problems, that is, binding of primers to nontarget DNA. Sequence similarity (discussed above) can cause mispriming, but sometimes primers will bind to DNA that lacks complementarity. This happens because most DNA samples contain a small amount of single-stranded DNA; at room temperature primers may bind nonspecifically to these fragments, and DNA polymerase may amplify them, even though it is relatively inactive at room temperature. You can prevent this by withholding a crucial part of the PCR mixture (e.g., magnesium) until the DNA sample has been heat denatured. One advantage of PCR is that it can be used to modify the DNA sequences that are amplified. Changing the DNA sequence is possible because PCR will still amplify a sequence even if one of the primers is slightly mismatched. The new DNA sequence is included in one of the primers, and as the cycles proceed, it will be incorporated into the amplified fragments. The aim in such a maneuver could be to alter the active site of an enzyme (protein engineering) in order to improve enzyme or protein function. More prosaically, it is often useful to introduce a sequence recognized by a specific restriction enzyme, so that a specific sequence can be cloned. The restriction site is usually added at the 5′ end of the primer, rather than the 3′ end, because the 3′ end is most responsible for the specificity of primer annealing. The addition of restriction sites is often an important step in the cloning of a PCR product, because it allows the introduction of sites that can be used to generate sticky ends. If successful, insertion of the amplified fragment into a plasmid vector is then straightforward. Longer sequences (up to 45 base pairs) can also be introduced at the 5′ end of a primer. Promoters and other regulatory elements are often added to an amplified segment in this way. Molecular biologists have also devised simple but

Gene Cloning and Production of Recombinant Proteins

65

lacZ’ MCS

lacI

amp

r

ori

FIGURE 3.11 General structure of the pUC family of plasmid vectors. (MCS = Multiple cloning site.)

elegant methods to introduce deletions or base substitutions of single base pairs or short sequences within a gene.

VII. ALTERNATIVE VECTORS: pUC The plasmid vector pBR322 is useful for explaining the basic strategy of gene cloning, but it is no longer a popular plasmid for most gene cloning procedures. The three major disadvantages to using pBR322 are (1) it is a low-copy-number plasmid; (2) the requirement of replica plating onto tetracycline plates is cumbersome and time consuming; and (3) most inserts in pBR322 are not strongly expressed (i.e., not much recombinant protein is produced). The pUC family of plasmid vectors (Figure 3.11) is superior to pBR322 in all three respects. The ori sequence is different and binds more readily to DNA polymerase, resulting in more frequent replication. This makes it easier and faster to obtain plasmid DNA (fewer bacteria are required to obtain a given amount of DNA). The promoter sequence is also different in pUC vectors. In pBR322 the tet promoter controls the transcription of inserted genes; in pUC vectors, the lacZ promoter has this function. The lacZ promoter binds to RNA polymerase more efficiently than the tet promoter, resulting in higher levels of mRNA formation and higher levels of translated protein. The final advantage of pUC vectors is in the method of selection and screening of transformed bacteria (Figure 3.12). Ampicillin treatment of bacteria transformed with pBR322 eliminates all bacteria that do not have the plasmid. Because pUC contains an ampicillin-resistant gene (amp), the same strategy works with pUC. In contrast to pBR322, though, identification of bacteria containing recombinant pUC

66

Introduction to Food Biotechnology

Desired gene Insert desired gene using the MCS

Transform E. coli

Plate on X-gal Subculture white colonies

FIGURE 3.12 Cloning a foreign gene into pUC. (MCS = Multiple cloning site.)

does not require replica plating. In pUC, restriction sites are clustered in a multiple cloning site (MCS), also known as a polylinker, that has been inserted at the end of the lacZ’ gene. The MCS consists of a series of restriction sites, allowing the user to pick the most appropriate restriction enzyme for digestion of the plasmid DNA and the foreign DNA that contains the desired gene. Successful recombination, then, results in incorporation of a foreign DNA sequence at the end of the lacZ’ sequence. This sequence contains only part of the lacZ gene, which codes for β-galactosidase, the well-characterized enzyme that is part of the lac operon in E. coli. When using pUC vectors, one uses a strain of E. coli that lacks the lacZ’ sequence but has the rest of the lacZ gene within its chromosome. When pUC is present in E. coli, the polypeptide expressed by the

Gene Cloning and Production of Recombinant Proteins

left arm

MCS

red+

gam+

67

MCS

right arm

FIGURE 3.13 Structure of λEMBL3. Only the left and right arms are required for lytic infection of E. coli. A multiple cloning site (MCS) is present next to each arm.

lacZ’ sequence in the plasmid interacts with the polypeptide produced by the chromosomal fragment in E. coli. This produces a functional β-galactosidase enzyme, which catalyzes the breakdown of lactose into glucose and galactose. Microbiologists have discovered that β-galactosidase also converts a synthetic carbohydrate (5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside, or X-gal) into a blue end product. Therefore, if a bacterium is transformed with pUC that has an intact lacZ’ gene, β-galactosidase will convert X-gal into a blue compound, and the colony will appear blue. However, if the bacterium is transformed with a recombinant pUC that has a foreign gene inserted onto the end of the lacZ gene, then the polypeptide expressed by lacZ’ will no longer interact successfully with the polypeptide expressed by the chromosomal fragment. As a result, the cell will lack β-galactosidase activity. Therefore, it will not convert X-gal into a blue compound; the colonies will remain white. This new protein is referred to as a fusion protein, because it can be visualized as the β-galactosidase protein with a foreign protein fused to it. Because the fusion protein lacks β-galactosidase activity, identification of recombinant clones is easy. The bacteria are diluted and plated on agar containing X-gal. White colonies are picked off and constitute a library of recombinant clones. Screening of a pUC library is similar to screening of a pBR322 library. The only advantage of using pUC is that it is easier to identify recombinant proteins because of their higher levels of expression in pUC. One final note about pUC vectors: they usually contain the lacI gene. This gene codes for the repressor protein that normally controls transcription of lacZ in E. coli. This protein prevents transcription of lacZ unless the inducer (lactose) is present. lacI allows pUC to regulate expression of the fusion protein. In practice, a synthetic inducer (isopropylthiogalactoside, or IPTG) is used instead of lactose.

VIII. ALTERNATIVE VECTORS: PHAGE LAMBDA A. LAMBDA CLONING VECTORS The Achilles heel of plasmid vectors such as pUC and pBR322 is their inability to efficiently incorporate sequences greater than 10 kb. For this reason, phage vectors, which can accommodate 20-kb segments of inserted DNA, are often used in the initial stages of gene cloning. λEMBL3 (Figure 3.13) is a commonly used example of a phage vector. (For a review of phage replication, see Chapter 2, Section IV.) λEMBL3 is very different from plasmid vectors. It is linear and it does not have an ori or antibiotic-resistant genes. Instead, it is made up of three regions: a left arm, a right arm, and a “stuffer” region. The stuffer region contains the red+ and gam+ genes. The crucial characteristic of the genes in the stuffer region is that they

68

Introduction to Food Biotechnology

cut vector with restriction enzyme Foreign DNA Left arm

a

stuffer

right arm

mix with foreign DNA fragments, and add DNA ligase

b c

in vitro packaging d

x

infect E. coli

FIGURE 3.14 Gene cloning using λEMBL3. Many different products (a, b, c, and d) result from ligation of λEMBL3 and foreign DNA. Product d is not packaged into phage particles because it is too short. Product c is not packaged because the right and left arms lack an intervening stuffer region. Products a and b are both packaged because they are the correct length and they contain right and left arms in the correct positions. The phage-containing product b, however, is unable to infect the P2 lysogen strain of E. coli because it contains the stuffer region.

are not required for the lytic cycle. In contrast, the left and right arms are essential for successful completion of the lytic cycle. Phage and plasmid vectors require a similar overall strategy. In both cases, the initial step is to cut the vector and DNA sample (containing the desired gene) with a restriction enzyme. λEMBL3 has been constructed so that it has MCSs at the border between the left arm and the stuffer region and between the right arm and the stuffer. Therefore, when restriction enzymes digest λEMBL3, the phage will be cleaved into three pieces: left arm, right arm, and stuffer. When the foreign DNA is cut with the same restriction enzyme and mixed with cut λEMBL3, a number of annealed products will appear (Figure 3.14). The only products of interest are strands made up of left arm ligated to foreign DNA ligated to right arm (product b in Figure 3.14). In these recombinant strands, the stuffer region has been replaced by a segment of foreign DNA. Our next challenge is to separate these recombinant DNA strands from the other annealed products. In fact, this is a very straightforward procedure. First, an in vitro packaging system is added. This is a solution containing all the viral proteins required for

Gene Cloning and Production of Recombinant Proteins

69

assembly of infectious λEMBL3 particles. When this solution is added to the mixture of annealed DNA, DNA segments will be assembled into viral particles. However, not all DNA segments will assemble. Only DNA that has an intact left and right arm at either end and is approximately 47 kb long will be incorporated into new virus particles. Therefore, many of the annealed products will not be incorporated, either because they lack the left or right arms, or because they are the incorrect length (see Figure 3.14). However, if the original phage DNA is still present (left arm + stuffer + right arm), then it will be repackaged into viral particles, because it is the correct length. These phage particles are eliminated in the next step: infection of a susceptible strain of E. coli. A specific strain of E. coli is used — P2 lysogen. λEMBL3 will complete its lytic cycle in a P2 lysogen only if it lacks the red and gam genes of the stuffer region. Therefore, any original phage particles will contain the stuffer, with its red and gam genes, and will not be able to replicate in the P2 lysogenic strain. Infection of E. coli will result in a library of recombinant viral particles that must be screened to find the virus that contains the desired gene. In principle, the screening is similar to the screening used to find the correct recombinant plasmid. The only difference is that virus particles from the library are diluted and plated onto a “lawn” of E. coli. When the diluted phage infect bacterial cells, they will progress through lytic infections, resulting in small cleared areas in the lawn (plaques). If the initial phage dilution was correct, then the plaques will be well separated, and it can be assumed that each plaque resulted from the initial infection of one bacterial cell by one virus. Each infected cell would then produce numerous new, identical virus particles, which would then infect neighboring cells. Eventually, enough neighboring cells would be infected and killed to produce a hole (plaque) in the lawn of bacteria. Each plaque represents a clone of recombinant λEMBL3. DNA probes can be used to screen these plaques for the correct clone. This is usually done by transferring DNA from the plaques onto a filter paper and then probing as described in Figure 3.8. Certain phage vectors (e.g., λgt11) lead to pronounced synthesis of introduced genes, which facilitates screening procedures based on detection of recombinant proteins. For example, antibodies specific to the desired protein can be used to identify plaques containing the protein.

B. COSMIDS Sometimes we need to clone segments of DNA that are more than 20 kb long. One alternative is to use cosmid vectors that can accommodate 40-kb segments. These are specialized plasmid vectors (Figure 3.15) that contain the two cos regions from lambda phage, as well as a restriction enzyme site within the cos regions, another restriction site, an antibiotic-resistant gene, and an ori allowing replication in E. coli. The cos sequences are the part of the left and right arms of the lambda phage that are required for packaging of phage DNA into new viral particles, as long as the DNA is the correct length (40 to 50 kb). Typical cosmids have an ori for replication in E. coli, as well as an antibiotic-resistant gene (ampr in Figure 3.15). This allows cosmids to replicate in host E. coli cells, similar to a plasmid vector,

70

Introduction to Food Biotechnology Restriction site b

cos

Restriction site a

cos

amp

r

ori

FIGURE 3.15 Structure of a typical cosmid vector.

and allows the use of antibiotics to eliminate cells that lack cosmids. Thus, cosmids have qualities of both lambda phage (cos) and plasmid (ori). Note also that our cosmid has two restriction sites that are cut by different enzymes (a and b). Restriction enzymes a and b cut the cosmid into two fragments (Figure 3.16). Both fragments are mixed with DNA fragments (~50 kb long) from the organism that has the desired gene. It is important that these genomic fragments are cut with restriction enzyme b. The cosmid and genome fragments are mixed and allowed to anneal. Some of the annealed (and ligated) segments will contain genomic fragments between two cosmid fragments. These products are the correct length to be packaged into lambda particles using in vitro packaging systems (see preceding section). The resultant viral particles will then inject the DNA into a host cell. However, the DNA contains only the cos regions from the left and right arms of lambda phage; consequently, they are unable to replicate new viral particles. At this stage the DNA recircularizes and behaves as a plasmid vector. Normal methods (e.g., cold CaCl2) of plasmid transformation of bacterial cells do not work with plasmids as large as cosmids. However, the phage injection system is able to introduce the large recombinant cosmids into host cells.

IX. ARTIFICIAL CHROMOSOMES One can easily sequence small sections within the genome of a eukaryote. However, understanding the order of genes along each chromosome can be an arduous task. This mapping process can be done more efficiently using vectors that can accommodate large sequences of DNA. Phage vectors and cosmid vectors are usually used, but artificial chromosomes are now available as an alternative. Artificial chromosomes were originally used in yeast. They contain centromere sequences that allow

Gene Cloning and Production of Recombinant Proteins

71

B

A

Cut with restriction enzymes A and B

+

cos

cos Add 50 kb fragments cut with restriction enzyme b

+

+

cos

cos Ligate

cos

cos

Package

Infect host cell

FIGURE 3.16 Cloning using a cosmid vector. The cosmid is cut with two restriction enzymes, generating two fragments. After ligation, the only segments that will be packaged into viral particles are those that are the correct length (because of the inclusion of the foreign insert) and have cos regions at either end.

a large segment of foreign DNA to function as a chromosome within a eukaryotic cell. Now they are also available for mammalian cells. Their latest application is as vectors for introducing DNA into animals, in addition to mapping genomes and locating specific genes within a set of chromosomes.

72

Introduction to Food Biotechnology

X. SUBCLONING AND VECTOR DIVERSITY Cloning is virtually never a one-step process. If the objective is to produce a recombinant protein in large quantities while keeping costs at a reasonable level, then it is unlikely that the initial cloning vector, and the organism “cloned into” will both be appropriate for large-scale production. In many cases we need to subclone. This is basically a repetition of the cloning process. For example, consider a situation in which we want to transfer a gene from Lactococcus lactis to Bacillus subtilis. It would theoretically be possible to set up a DNA library directly in B. subtilis, if an appropriate plasmid or phage vector were available. However, it would be easier to set up the library in E. coli, screen the library, and isolate the desired clone. The desired gene could then be cut out of the cloning vector and subcloned into a plasmid vector that is able to replicate in B. subtilis. This is an example of a shuttle vector (it is used to “shuttle” the cloned gene from E. coli to B. subtilis). Subcloning can have other aims. Often, the vector used in the primary cloning procedure does not result in efficient protein expression; subcloning into an expression vector, with strong promoters and appropriate signal sequences (see Chapter 2, Section V), will often increase the amount of secreted heterologous protein. Vectors with signal sequences are also known as secretion vectors. Finally, before the advent of PCR, which has greatly simplified DNA sequencing, cloned genes were usually subcloned into sequencing vectors, which have characteristics appropriate for DNA sequencing (e.g., the generation of single-stranded DNA).

XI. RECOMBINANT CHYMOSIN A. CHYMOSIN

AND

CHEESE MAKING

Chymosin has a long history of use in food production and is essential to cheese making practices throughout the world. When added to acidified milk, chymosin increases the rate and extent of coagulation of casein, the predominant protein in milk (see Figure 1.3). This coagulated protein (the curd) is then treated in various ways to make cheese. Typically, the liquid whey is drained or pressed from the curd, which is then left for variable amounts of time to ripen. Traditionally, rennet has been the source of chymosin. Rennet is taken from the stomachs of calves after slaughter. Chymosin is a protease that is produced by epithelial cells in the fourth stomach of calves. The main problem with the use of chymosin from calves is that the supply is unstable and is affected by global demand for veal. Purity is an additional problem; different batches and chymosin from different suppliers are not always of similar purity and activity. Purity is important because animal rennet is contaminated with variable levels of bovine pepsin, which causes undesirable proteolysis and unwanted flavors if present during curd formation. Chymosin is a complex protein. The protein that is translated from the mRNA transcript in calf epithelial cells is initially inactive (Figure 3.17). This form is preprochymosin. The “pre” part of the protein is a signal sequence of amino acids

Gene Cloning and Production of Recombinant Proteins

nucleus signal sequence preprochymosin mRNA ribosome

73

Stomach lumen

prochymosin autocatalytic cleavage chymosin Intestinal epithelial cell

FIGURE 3.17 Formation of chymosin in the lumen of the fourth stomach of bovine calves. The mRNA is exported from the nucleus and is translated into preprochymosin. The signal sequence is removed and prochymosin is secreted from the cell. The low pH of the stomach lumen causes an autocatalytic cleavage of prochymosin, yielding chymosin, the active form of the enzyme.

that informs bovine intestinal epithelial cells that preprochymosin is a secreted protein. This signal sequence is cleaved from the protein as it passes across the plasma membrane into the intestinal lumen. Therefore, prochymosin is the form of chymosin that is secreted into the lumen. Interestingly, it has autocatalytic activity and can “cut itself up.” However, this cleavage occurs only in acidic conditions. When prochymosin is secreted into the acidic environment of the fourth stomach, part of the protein is cleaved and the remaining protein (chymosin) is an active protease. In cheese production, chymosin is added to milk shortly after the starter culture. Lactic acid bacteria in the starter ferment carbohydrates in the milk and produce lactic acid as an end product. This results in increased acidification of the milk. In the absence of chymosin, decreased pH causes coagulation of casein, the predominant protein in milk. However, a soft curd results, similar to soft cottage cheese. To get a firm curd, chymosin must be added. Chymosin causes increased coagulation because it cleaves casein at points in the amino acid chain that are covalently attached to carbohydrate chains (Figure 3.18). Intact casein exists in aqueous solution as a micelle, and aggregation of micelles is prevented by electrostatic repulsion between carbohydrate chains in adjacent micelles. However, when chymosin removes these carbohydrates from the micelles, nothing prevents aggregation, and a firm curd results. The mechanism of action of chymosin is very specific, so it is essential that the recombinant protein be identical in structure to the protein produced in calf stomachs. Various microbial proteases can be substituted for chymosin, but they usually yield inferior cheese because of excessive proteolysis. A better alternative would be to transfer the bovine gene for chymosin to a microorganism that can be grown on a large scale at low cost. Several research groups attempted to do this in the early 1980s.

74

Introduction to Food Biotechnology

Hydrophobic core Casein protein Carbohydrate chain chymosin

Casein micelle

Aggregation of micelles

FIGURE 3.18 Mode of action of chymosin on casein micelles.

B. INCLUSION BODIES Initial efforts to clone chymosin into E. coli were partially successful. mRNA was isolated from calf intestinal epithelial cells, and cDNA libraries were constructed using plasmid vectors such as pBR322. These libraries were screened using a variety of strategies (one research group used DNA probes designed from part of the chymosin amino acid sequence). Recombinant clones of E. coli were isolated, and the presence of the prochymosin gene was confirmed by sequencing of the inserted DNA. However, when the chymosin gene was subcloned into an expression vector with strong promoters, problems arose. The protein was produced in large quantities (5% of total cellular protein), but it was not secreted from cells; instead, it accumulated within cells in clumps of denatured protein. Such clumps, referred to as inclusion bodies, commonly occur when foreign proteins are expressed at high levels in E. coli and other bacteria. This is frequently a problem when transferring genes from eukaryotes into E. coli. The intracellular environment (e.g., pH, redox potential) is very different in bacteria than in eukaryotes. This often causes aggregation of eukaryotic proteins. Also, many eukaryotic proteins undergo processing in the endoplasmic reticulum and the Golgi apparatus, which form the internal membrane system of eukaryotic cells (Chapter 2, Section V). These membranous compartments are chemically different from the cytoplasm. For example, the oxidizing environment of the endoplasmic reticulum allows disulfide bond formation in proteins that would be unlikely to occur in the more reducing conditions of the cytoplasm. Furthermore, eukaryotes often have molecular chaperones that help ensure that proteins are folded correctly. In prokaryotes, the chaperones have different structure, sometimes leading to incorrect folding of proteins, which may lead to exposure of hydrophobic residues to the aqueous solution. The interaction of such exposed regions among different protein molecules results in extensive aggregation and denaturation of the protein.

Gene Cloning and Production of Recombinant Proteins

75

2 micron segment

amp

leu

r

ori

FIGURE 3.19 Structure of YEp (yeast episomal plasmid).

Inclusion bodies can be purified from E. coli and renatured, but this process is usually expensive. There are better alternatives. In the case of chymosin, the most successful approach has been to subclone the prochymosin gene into fungal hosts. Because fungi are eukaryotes, inclusion bodies are much less of a problem.

C. RECOMBINANT PRODUCTION

BY

YEAST

One of the advantages of using yeast (unicellular fungi) in cloning programs is that there are plasmids that are able to replicate in yeast and can be used as vectors. One such plasmid is yeast episomal plasmid (YEp). Episomes are segments of DNA that are able to integrate into host chromosomes or remain extrachromosomal. Under certain conditions YEp is maintained at a high copy number (30 to 50 copies per cell) in yeast cells and is therefore not diluted out as the cells divide by mitosis. In comparison to bacteria, it is more difficult to transform yeast with DNA. However, several methods work; electroporation is particularly effective with yeast sphaeroplasts (cells whose walls have been enzymatically removed). This method requires the application of high voltage to a suspension of yeast cells. Small pores transiently form in yeast membranes, allowing the uptake of DNA into the yeast cells. Biotechnologists have constructed YEp shuttle vectors that can replicate in both E. coli and the yeast Saccharomyces cereviseae (Figure 3.19). After cloning the prochymosin gene in E. coli, they subcloned the gene into a YEp vector and incorporated it via transformation into S. cereviseae. Large amounts of protein were produced, but unfortunately, it was in an insoluble, aggregated form. This occurred because prochymosin was not secreted from the yeast cells. The original form of the gene in cows never accumulates intracellularly, because it is

76

Introduction to Food Biotechnology

rapidly secreted from intestinal epithelial cells. Therefore, it seemed sensible to try to mimic this, by persuading recombinant yeast cells to secrete recombinant prochymosin. So, another subcloning step was required. This time, the prochymosin gene was subcloned into a secretion vector. Such vectors already have a signal sequence recognized by S. cereviseae. Any foreign genes inserted in front of this sequence have a much better chance of being secreted from host yeast cells. The secretion vector was partially effective. A large proportion of the translated protein was secreted from the yeast cells. Unfortunately, the yield of recombinant protein was too low; prochymosin production was not cost effective. The low yield problem is intrinsic to S. cereviseae, a fungus that normally does not secrete large amounts of proteins. The solution to the prochymosin conundrum finally came with a final subcloning step that allowed the transformation of a different yeast — Kluyveromyces lactis. This yeast normally secretes large amounts of a variety of enzymes, and has been used for a number of years to produce commercially sold lactase. When a recombinant plasmid containing the prochymosin gene was introduced into K. lactis, sufficient amounts of recombinant prochymosin were synthesized and secreted into the growth medium. Chymosin could then be easily collected, purified, and sold to cheese makers. Recombinant chymosin has also been produced successfully in filamentous fungi (e.g., Aspergillus nidulans). The chymosin story is typical of many attempts to produce recombinant proteins in large quantities. Initial cloning efforts are rarely completely successful, and usually a number of subcloning steps and genetic modifications are necessary to produce recombinant proteins at a reasonable cost. What about safety assessment of chymosin? The U.S. Food and Drug Administration (FDA) considers recombinant chymosin to be sufficiently similar to natural chymosin. Consequently, toxicity assessment was not required, and permission was granted to use recombinant chymosin in the production of cheese. Several other factors influenced the FDA’s decision; for example, purification of chymosin resulted in destruction of the microbe that produced the chymosin and efficiently removed impurities associated with large-scale culture of the fungus. Chymosin can also be used for cheese production in the European Union, Canada, and many other countries.

XII. RECOMBINANT BOVINE GROWTH HORMONE Before the advent of rBGH, it was not possible to use BGH to boost milk production in cows. This is not because of a lack of understanding of the potent biological effects of injected BGH on cows. As early as the 1940s, animal scientists knew that injections of this hormone would increase the rate of milk production in cows. However, at that time the only source of BGH was from dead cows, and the cost of obtaining useful amounts was prohibitive. Consequently, research on the potential uses of BGH stalled, until the development of recombinant BGH in the early 1980s. This illustrates again the major advantage of recombinant proteins — they are inexpensive. Once the gene for BGH was transferred to E. coli, it became economically feasible to grow large amounts of the recombinant bacterium, purify the hormone, and sell it at a price that is cost effective for dairy farmers.

Gene Cloning and Production of Recombinant Proteins

77

rBGH is an excellent example of the nontechnical problems that a biotechnological company can face when attempting to introduce a biotechnological product into the food supply. Attempts to introduce rBGH to markets throughout the world have met with extensive opposition. This is surprising, because of the lack of controversy surrounding many other recombinant proteins. For example, both recombinant insulin and recombinant chymosin have been successfully sold throughout the world, with very little controversy. Opponents of rBGH fear its effects on cows and on humans who drink cow’s milk. In the U.S., the FDA was principally responsible for assessing the safety of rBGH, based on studies by the company (Monsanto) that developed rBGH, as well as independent studies. When there is a question of potential toxic effects of a recombinant protein, it is necessary to test the toxicity of the protein using animal models. Animals are exposed to the protein and closely studied for adverse changes. Toxicological studies did not detect any adverse changes in animal health after administration of milk from rBGH-treated cows. However, many opponents to rBGH have little faith in standard toxicological studies, because of the inability of such studies to detect adverse changes over a long period of consumption of the product or compound. Because of these problems in assessing the safety of long-term consumption of new compounds, it is important to carefully consider differences between the new product and analogous, currently used products. In the case of rBGH, then, the central question is whether or not milk from rBGH-treated cows is different from milk from untreated cows. It is virtually impossible to state unequivocally that rBGH milk is identical to normal milk; milk is a complex suspension and solution that contains a myriad of different biochemicals. For this reason, the FDA felt that it was feasible to examine only specific compounds, to see if levels were affected by rBGH treatment. They found that levels of BGH were similar in both kinds of milk, but there was a statistically significant increase in the amount of insulin-like growth factor (IGF-1) in milk from rBGH-treated cows. IGF-1 is a protein hormone that is influenced by growth hormones such as BGH; typically, increased levels of growth hormones lead to increased levels of IGF-1. The controversy surrounding IGF-1 is largely related to its potential role in tumor formation. One of the principal actions of IGF-1 is to increase the rate of cell proliferation, which also occurs during tumor formation. To determine if a byproduct such as IGF is potentially dangerous to human consumers of milk, the following questions must be considered: 1. Would IGF survive its voyage from the oral cavity to the intestine? To do so, a compound must not be affected by transit through the stomach, with its low pH. 2. Is IGF absorbed by the intestine into the body, and into the circulation? 3. Does bovine IGF exert any effects on cells or processes of the human body? Would a bovine hormone such as IGF bind to receptors in human cells? If the answer to any of these questions is “yes,” the potential risk to humans increases.

78

Introduction to Food Biotechnology

The FDA decided that the available scientific evidence indicated that increased levels of IGF-1 in rBGH-treated milk were not dangerous. The primary reason for this decision was that it appeared unlikely that ingested IGF-1 would remain intact after passage through the stomach. It also determined that the amount of IGF-1 ingested would be insignificant, partly because of the relatively large amount of IGF-1 present in normal human saliva. The ability of human intestinal epithelial cells to absorb IGF-1 and the ability of bovine IGF-1 to exert effects on human cells were considered irrelevant, because of the stomach and saliva considerations. The FDA’s approval of the sale of rBGH-treated milk has not eliminated controversy over this product. One of the principal points of controversy surrounds the ability of many pH-sensitive microbes (e.g., Salmonella typhimurium) to survive passage through the stomach under certain conditions (e.g., protection by lipids). Is it possible for IGF-1 to survive passage through the stomach? The level of controversy surrounding IGF-1 and rBGH has now abated in the U.S., but most countries throughout the world (including the European Union) still do not allow the sale of rBGH-treated milk. In some countries (e.g., Canada) rBGH did not receive approval because of studies indicating adverse effects on cow health. Several studies have found that administration of rBGH results in increased rates of mastitis (infection of the udders) and other health problems. These problems are probably linked to increased milk production, and not directly to rBGH treatment. It may seem odd that BGH has generated so much controversy. Part of the problem may lie in the basic nature of BGH; many people are offended by the idea of “supercharging” cows with growth hormones to increase milk supply. Thus, the fact that rBGH is a product of recombinant DNA technology may be less important than the fact that it is a potent hormone. The use of “natural” BGH derived from carcasses would likely be considered equally objectionable to many people. However, a significant number of consumers appear to be offended by the “unnatural” aspects of rBGH (i.e., it is a mammalian protein produced by a bacterium). Interestingly, rBGH is safer than natural BGH, because natural BGH could be contaminated with viruses or the causative agent of bovine spongiform encephalopathy (mad cow disease). Consumer benefit is another important issue. The use of rBGH does not directly benefit the consumer; it does not increase the quality of milk and it is unlikely to result in decreased milk prices. This adds to consumer resentment. In contrast, the use of recombinant chymosin is simply a change in the source of an enzyme that has always been used to make cheese. Thus it is less likely to be seen as a technological intrusion into a traditional food. Also, the use of recombinant chymosin may make cheese more palatable to lacto-vegetarians, who are likely to be offended by the use of chymosin derived from slaughtered calves. Also, Muslims and Jews may prefer to use cheese made from recombinant chymosin to ensure that they are not consuming cheese made using porcine (pig-derived) chymosin. There is an important lesson here for food biotechnologists — the general public is particularly sensitive to products that contain elements that are central to human metabolism, such as growth hormones. Other heterologous proteins, such as recombinant chymosin or recombinant amylases (used in starch processing), have largely

Gene Cloning and Production of Recombinant Proteins

79

been ignored by antibiotechnology activist groups, and generate very little controversy. Another lesson: if a biotechnologically derived food or process can be seen to benefit the consumer, it is much more likely to have a seamless introduction into the marketplace. Evidently, the potential public reaction to a biotechnological product should be a primary point for biotechnologists to ponder in the early stages of new product development.

RECOMMENDED READINGS 1. Sambrook, J. and Russell, D., Molecular Cloning: A Laboratory Manual, 3rd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 2001. 2. Rapley, R., Molecular analysis and amplification, in Molecular Biology and Biotechnology, 4th ed., Walker, J. M. and Rapley, R., Eds., Royal Society of Chemistry, Cambridge, 2000, chap. 2. 3. Rapley, R., Recombinant DNA technology, in Molecular Biology and Biotechnology, 4th ed., Walker, J. M. and Rapley, R., Eds., Royal Society of Chemistry, Cambridge, 2000, chap. 3. 4. Hengen, P. N., Methods and reagents — fidelity of DNA polymerases for PCR, Trends Biochem. Sci., 20, 324, 1995. 5. McPherson, M. J., Møller, S. G., Beynon, R., and Howe, C., Pcr (Basics: From Background to Bench), Springer-Verlag, Heidelberg, 2000. 6. Newton, C. R. and Graham, A., PCR, 2nd ed., Springer-Verlag, New York, 1997. 7. Harris, E. and Kadir, N., A Low-Cost Approach to PCR: Appropriate Transfer of Biomolecular Techniques, Oxford University Press, Oxford, 1998. 8. Glick, B. R. and Pasternak, J. J., Molecular Biotechnology: Principles and Applications of Recombinant DNA, ASM Press, Washington, D.C., 1994. 9. Glazer, A.N. and Nikaido, H., Microbial Biotechnology: Fundamentals of Applied Microbiology, Freeman, New York, 1995. 10. Ward, M., Chymosin production in Aspergillus, in Molecular Industrial Mycology: Systems and Applications for Filamentous Fungi, Leong, S. A. and Berka, R. M., Eds., Marcel Dekker, New York, 1991, chap. 4. 11. Beppu, T., Production of chymosin (rennin) by recombinant DNA technology, in Recombinant DNA and Bacterial Fermentation, CRC Press, Boca Raton, FL, 1988, chap. 2 12. Juskevich, J. C. and Guyer, C. G., Bovine growth hormone: human food safety evaluation, Science, 249, 875, 1990. 13. Bonneau, M. and Laarveld, B., Biotechnology in animal nutrition, physiology and health, Livestock Prod. Sci., 59, 223, 1999.

4

Plant Biotechnology I. OVERVIEW

The plant biotechnology field is of prime importance to anyone studying food production, food science, or nutrition. Some new plant biotechnologies (e.g., the use of plant cell culture) have gained public acceptance and are used throughout the world; others, particularly those surrounding the development of transgenic plants, are controversial and, in many parts of the world, have not been accepted by the public. The central aim of this chapter is to explore the kinds of biotechnologies that affect crop production, the processing of food plants, and the composition of food plants. An emphasis is placed on transgenic plants because they have great potential to improve a diverse array of problems associated with food production and processing as well as health-promoting aspects of food. Also, transgenic plants are clearly the most controversial application of food biotechnology, and consequently have the potential to disrupt matters ranging from the global (international trade of food commodities) to the local (changed patterns of food consumption due to consumer concerns about the safety of transgenic crops). This chapter explores plant cell and tissue culture techniques because of their central importance in traditional plant breeding and the production of transgenic plants. Plant cell culture allows the rapid production of genetically identical clones of valuable plants; this is exploited on a day-to-day basis in the cultivation of certain crops (e.g., potatoes) and is a valuable tool for plant breeders. Techniques such as embryo rescue and protoplast fusion allow plant breeders to expand the range of plants that can participate in breeding programs to plants that belong to different species or genera. Normally, breeders must choose parents from varieties of the crop species, restricting the range of available genetic material. With in vitro (culture of cells or tissues outside the plant environment) techniques, though, breeders can transfer valuable genes between species and sometimes between genera. Plant cell culture is also a potentially useful technology in the context of food processing; a number of compounds used to influence food flavor and aroma (e.g., saffron) are expensive to obtain from intact plants. An alternative is to grow cells in culture and isolate the desired compound from the cells or the culture medium. To date, though, this technology is not commercially popular because of the relatively high cost of growing plant cells. Cell culture techniques are also required for the production of transgenic plants. Most transgenic plants that have been released to growers have been aimed directly at helping the grower, by making it cheaper or easier to grow commodities such as corn, soybeans, and canola (low-erucic-acid rapeseed). However, a number of transgenic tomato, canola, and rice varieties have been developed with food processing 81

82

Introduction to Food Biotechnology

or nutritional objectives. Most available tomato transgenics have slower ripening, or less softening accompanying ripening, than conventional tomatoes. One of the transgenic canola cultivars that has been released has an altered pattern of production of polyunsaturated fatty acids (PUFAs). In this case, the new PUFA has nonfood industrial applications, but the same technology is currently being tested in the context of altering PUFA production to increase levels of health-promoting PUFAs (e.g., linolenic acid). Published studies have also demonstrated that rice and other crops can be modified to have increased levels of provitamin A or vitamin E. Although these transgenic cultivars have not yet been cleared for commercial release, they demonstrate the power of transgenic technology to alter important properties of food. The reaction to transgenic plants is not always enthusiastic, though. Transgenic plants have received an enormous amount of attention from the media and public. This attention has been triggered by vigorous action of mainstream and radical activist groups, both at international and grassroots levels. Because of this public interest, it is vital that plant scientists, food scientists, nutritionists, and biotechnologists understand the technology that is used to develop transgenic plants. Because of the public backlash against transgenic plants, many biotechnology companies have scaled down their transgenic plant research and development programs. However, several large biotechnology companies have continued to release transgenic cultivars, perhaps because they have the financial resources to risk public rejection of transgenic food. Commercialization of transgenic crops requires many years of basic research and testing, followed by exhaustive (from the point of view of biotechnology companies) submissions to regulatory agencies such as (in the U.S., for example) the Food and Drug Administration (FDA) and the Environmental Protection Agency (EPA). To minimize risk, companies attempt to commercialize transgenic crops that will be attractive to a large number of growers, usually because of decreased pesticide costs. Thus there has been little direct benefit to consumers, which is one of the reasons for the low level of public acceptance of transgenic crops in many parts of the world. It is predicted that the next generation of transgenic cultivars will offer consumers great benefits (e.g., increased levels of nutraceuticals in food), making transgenic technology more palatable to the public. However, the financial risks associated with transgenic plant commercialization have made it difficult for biotechnology companies to take the risk of developing cultivars aimed at the consumer market. This has led to a great deal of dissatisfaction among biotechnologists involved in the development of transgenic plants. They feel that antibiotechnology activists have unfairly targeted their technology, and that the risks of transgenic plant technology have been greatly overstated. On the other hand, antibiotechnology activist groups, as well as a number of consumer lobby groups, feel that transgenic crops have been inadequately assessed for health and environmental risks, and that the release of transgenic seeds to the farming community, especially in the U.S. and Canada, constitutes an unacceptable lapse in governmental regulation. It is unlikely that this issue will be resolved soon to the satisfaction of either camp. It is unlikely that transgenic technology will disappear, and it is unlikely that antibiotechnology groups will ever be satisfied with the processes used to regulate commercialization of transgenic crops. Some of the environmental concerns

Plant Biotechnology

83 Shoot apical meristem

Axillary meristem Vascular cambium

Root apical meristem

FIGURE 4.1 Points of growth in plants. Axillary meristems represent potential branching points of the stem (shoot), and the vascular cambium is responsible for increasing the girth of a stem or root. The cork cambium is not shown in this diagram, but it is required for the production of bark in woody plants. Vascular and cork cambia are lateral meristems.

associated with transgenic plants will be discussed in this chapter (Sections IV.A and IV.C). Health concerns are discussed in Chapter 9. In this chapter, plant cell culture and several of its applications will be discussed first. This will be followed by an explanation of the techniques that are used to directly transfer genes into crop plants, then examination of the range of realized and potential applications of transgenic plant technology.

II. PLANT CELL AND TISSUE CULTURE A. CONTROL

OF

PLANT GROWTH

To understand how plant cells are manipulated in vitro, we must understand how plants grow and how such growth is controlled. Plants, like all multicellular organisms, grow through the combined action of two processes: cell division and the expansion of individual cells. In most animals, a well-defined pattern of cell growth and division occurs early in life, leading to an adult organism that no longer increases in size. However, in many plants (especially trees), the full extent of growth is more variable, and growth occurs throughout the plant’s lifetime. This growth is carefully controlled by the plant, and it only occurs in meristems (Figure 4.1), which are most easily seen at the tips of roots and stems (apical meristems). Less obvious are axillary meristems, dormant buds that are located in the axils of leaves. An

84

Introduction to Food Biotechnology

axillary meristem, when given the appropriate hormonal stimulation, develops into a new growing apical meristem; this is how branches form. Lateral meristems, consisting of the vascular cambium and the cork cambium, are found in the interior of woody stems and roots. These meristems produce woody tissue and bark, acting to increase the girth of a plant to support continued upward growth of the apical meristems. If this did not occur, the stem would be unable to support the weight of the new growth occurring at the tips of branches. So, how is meristem growth controlled? Evidently, control is essential; uncontrolled growth and subsequent tumor formation is damaging and wasteful for plants, just as it is for animals. In plants, control of growth occurs via plant hormones. The most well-characterized hormones are the auxins, gibberellins, and cytokinins. They are a source of great confusion because of the myriad of effects they have on plant cells. Effects vary within plants; for example, auxin will likely inhibit cells in an axillary meristem (i.e., the cell will not divide or expand), whereas it would stimulate cells to divide in the vascular cambium. To complicate matters further, cells from different plant species have different patterns of response to plant hormones. Fortunately, we can make some general comments on these classes of hormones: • Auxins tend to stimulate the expansion of cells, by triggering the loosening of cell walls. This allows turgor pressure to increase the size of cells (similar to the increase in size of a balloon when water flows into the balloon). • Gibberellins tend to stimulate both cell division and cell expansion. This is particularly evident when gibberellins are applied to germinating seedlings. • Cytokinins tend to stimulate cell division. These hormones are often produced in apical meristems and cause cells in the meristem to divide.

B. THE IN VITRO LIFE CYCLE These hormones are obviously interesting to plant scientists, but why are they important to food biotechnologists? The answer lies in the use of plant hormones to stimulate the growth of plant cells in vitro. To illustrate this idea (Figure 4.2), consider a tobacco plant growing in soil in a pot. If you remove a tobacco leaf and place it in soil in another pot, it will almost certainly die and quickly decompose. However, if you use a hole-puncher to cut out pieces of a leaf (explants) and place the pieces on an agar medium containing mineral nutrients, a source of carbohydrates (usually sucrose), and plant hormones, then growth will occur. The type of growth can be manipulated by changing the types of hormones present and the concentration of each hormone. For example, callus formation is sometimes useful. A callus is a disorganized mass of dividing, undifferentiated plant cells. Often, a relatively high concentration of auxin in the agar medium stimulates callus formation. Once a callus has formed, various paths are possible. The clumps of cells can be broken up in a liquid medium, forming a cell suspension. Such cell suspensions can be grown indefinitely, if nutrients and the correct mixture of hormones are

Plant Biotechnology

85

b. callus formation a. remove explant

c. cell suspension

d. callus formation e. shoot formation

FIGURE 4.2 Tissue culture of a tobacco plant. In (a), a leaf explant is transferred to agar media containing a specific combination of plant hormones. This combination of hormones induces the formation of callus tissue (b). Callus tissue can be converted into cell suspensions in liquid media (c). Transfer of callus or cell suspensions to agar media with another specific combination of plant hormones results in induction of the growth of intact plantlets (d). Plantlets can be transferred into soil (e), and then into the field, if desired.

supplied. As long as the suspension is gently shaken, the cells will remain suspended and callus tissue will not form. Cell suspensions can in turn be returned to solid agar media, leading again to callus formation. Pieces of callus can then be transferred to fresh media, allowing the culture of callus indefinitely. Often, though, it is desirable to manipulate callus cultures so that intact plants (“plantlets”) grow from the callus. Specific changes in the concentrations of hormones in the agar medium stimulate plantlet formation. For example, transfer of calli to agar containing a high concentration of cytokinins and a low concentration of auxin may result in the regeneration of shoots. Once shoots have formed, further manipulation of the hormones may be necessary to induce root development. Once shoots and roots are present, the plantlets can be transferred to soil. This transfer is often very traumatic, and plantlets usually have to be carefully nurtured in soil to prevent an early death. This conversion from cells to callus to intact plants can start with single plant cells, helping to ensure that the resulting plant is genetically uniform. In comparison to animals, manipulation of plant cells is relatively easy. The main reason for this is that totipotent cells are easier to obtain in plants. A totipotent cell has the potential to differentiate into any of the cells present in the adult plant. In animals, totipotency is found only in the early stage of embryo formation. Plants, in contrast, always have totipotent cells in their meristems, especially in shoot apical and axillary meristems. Thus, adult plants retain small populations of embryonic cells that continue to divide and differentiate throughout the life of the plant. In some plants, (e.g., tobacco), totipotent cells can also be obtained from nonmeristematic cells (e.g., photosynthetic cells in the leaf). In other plants (e.g., many cereals), it is difficult to isolate totipotent cells; they may be found only in embryonic tissue. This is the major reason for the lag in development of transgenic barley, wheat, and rice, despite the great economic importance of these crops.

86

Introduction to Food Biotechnology

C. MICROPROPAGATION This ability to circumvent the normal life cycle of plants is useful. It is often desirable to produce large numbers of plants that are genetically identical. Such plants may have characteristics that lead to faster growth, higher yield, or other useful agricultural characteristics. Alternatively, they may have superior processing characteristics; for example, a particular potato variety may have low levels of sucrose in the tubers and thus be better suited for the production of fried, frozen potato chips. This variety of potato could be propagated in the usual way, by growing plants in the field, harvesting tubers, and selling them to farmers as seed potatoes. Alternatively, the variety could be micropropagated on agar media, potentially producing hundreds of thousands of genetically identical plantlets. These plantlets could then be sold to farmers as an alternative to seed potatoes. One advantage of micropropagation is that it is easier to ensure that the propagated plants are free of viruses and other pathogens. This can be a problem for plants that are propagated through vegetative means (i.e., not by seeds), such as potatoes. It is difficult to ensure that potato tubers grown conventionally are virus free, but virus-free plants can readily be obtained using tissue culture techniques. One of the best strategies for producing virus-free plants is to establish cell cultures from cells taken from apical meristems. These cells are not fed directly by the plant’s vascular system, and, because many viruses spread through the vascular tissues, meristems remain virus free. Micropropagation is particularly useful for crops such as potato and banana, which are not propagated by seed, and is less popular for seed-propagated crops such as cereals and legumes. However, micropropagation is frequently an essential tool in plant breeding programs aimed at improving seed-propagated crops. Micropropagation is not risk free. Biotechnologists try to avoid callus formation when micropropagating plants. Cells in callus culture are genetically unstable; mutations frequently occur that may damage the cell or result in the loss of useful characteristics. This is due to somaclonal variation (see Section II.D.3 of this chapter). Somaclonal variation can be devastating if desirable characteristics are lost or if undesirable characteristics are gained. For this reason, micropropagation systems are designed so that they minimize the time spent in callus culture. Large-scale micropropagation requires the monitoring of propagated plants, to ensure that somaclonal variation does not gradually change their genetic constitution.

D. PLANT CELL CULTURE

AND

TRADITIONAL PLANT BREEDING

1. Plant Breeding Basics The improvement of crop plants has traditionally been the realm of the plant breeder. Most plant improvement is still achieved through traditional plant breeding techniques; cell culture is often an important component of plant breeding programs. What is traditional plant breeding? It is important for food biotechnologists to understand the answer to this question, because plant breeding is the method used to develop most new food crops. Traditional breeding is preferred to a transgenic

Plant Biotechnology

87

approach when numerous genes control the target for improvement. Yield is a good example — the amount of harvested product yielded by a plant is affected by (1) genes that affect distribution of photosynthesis-derived carbohydrates, (2) genes controlling plant architecture (e.g., how tall the plant is or how many branches it has), and (3) genes relating to the ability of the plant to tolerate stresses such as dry soil. Recombinant DNA technology is very effective at inserting single genes into plant cells, but as the number of inserted genes increases, the difficulty quickly increases. Consequently, the transgenic plants that have been released to date consist mainly of one- or two-gene insertions. As transgenic technology evolves, and as we learn more about the genetics of plant development and function, people will develop transgenic plants with complex characters. Actually, this is already in progress; biotechnologists have successfully developed transgenic rice varieties with increased yield, as well as wheat varieties with increased tolerance to drought stress. In the future, transgenic technology will be used more often to improve consumer-important characteristics such as appearance, flavor, and functionality. However, traditional plant breeding is likely to continue to be the method of choice for many plant improvement programs. For this reason, it is essential to understand the primary tool for improvement of plant foods. Traditional plant breeding (Figure 4.3) requires first and foremost the identification of desired characteristics. For example, an agronomist (one who is concerned with the behavior of field-grown plants) may wish to develop a variety of wheat that stays compact and flowers quickly, so that the plant invests most of its energy in reproduction (i.e., harvestable grains). Alternatively, a company that sells frozen vegetables may wish to have a variety of carrot that does not deteriorate in quality during storage. A competent plant breeder can grant both of these wishes, although it might take 5 to 10 years. Once the desired characteristic has been identified, the breeder starts with a cultivar (varieties are called cultivars if they are being actively cultivated by producers) that is currently successful with growers and that will be useful to the intended market. For example, if a breeder wants to develop a variety that will be used by greenhouse tomato growers in the American midwest, then s/he will select a cultivar that is currently popular with these growers. Plant breeders usually select cultivars that will appeal to a large number of producers. The next task is to identify cultivars or varieties, or, in some cases, related wild plants that have the desired new character. For example, a common target of plant breeding programs is enhanced resistance to fungal pathogens. Cultivar X is the chosen cultivar that performs well but lacks resistance to a particular fungus. Cultivar Y lacks many of the valuable characteristics of cultivar X, but it is resistant to the fungus. The plant breeder will then cross cultivars X and Y, by taking pollen from one cultivar to pollinate the other. Some of the offspring (the F1 generation) from this cross should have enhanced resistance to the fungus, while retaining most of the good characteristics of cultivar X. In terms of classical genetics, the phenotype (resistance) of cultivar X has changed, because of alterations in the genotype (genetic structure) of cultivar X. In the second round of crosses, the breeder selects plants from the F1 generation that have enhanced levels of fungal resistance.

88

Introduction to Food Biotechnology Parent variety

Excellent taste, low carotene

Select source of desired gene(s)

Poor taste, high carotene Backcross with original parent, if valuable traits have been lost

Make first cross: remove pollen from one variety, and fertilize flower from the other variety

Assess carotene content of the offspring of the first cross

Make second cross with best offspring Assess carotene content of the offspring of the second cross

FIGURE 4.3 The use of traditional plant breeding techniques to develop a carrot variety with increased levels of β-carotene. The parent variety has a number of valuable traits (e.g., sweet taste), but has low levels of β-carotene. The breeder introduces genes from a carrot variety that has high levels of β-carotene, through careful pollination. Seeds are grown from this cross, and the plants are tested for β-carotene and other traits. The second cross is made between plants with enhanced levels of β-carotene. This process is repeated until the desired level of β-carotene is achieved. However, the resulting plants may lack some of the original valuable characteristics of the parent variety (e.g., they may have low levels of sucrose). Backcrossing with the parent variety is then required to restore the missing traits. Once this is achieved, the breeder attempts to breed a “true-breeding” line (variety) that does not vary in its important traits. Finally, registration is sought; this requires a submission to a regulatory agency of test results demonstrating that the new variety performs acceptably under a variety of growing conditions in a broad geographic range, and that the new variety offers a substantial improvement over previous varieties. If successful, the breeder will then have “plant breeder’s rights,” which in most countries offer a degree of intellectual property protection that is less powerful than patent protection but still valuable to the breeder.

Plant Biotechnology

89

The biggest problem with this approach is that after several generations the plants will likely have lost many of the valuable characteristics of cultivar X. This occurs because of the nature of meiosis, the process that leads to pollen and ovule formation. Each pollen or ovule has only one half of the chromosomal complement of the parent plant. So in a sense, the breeder has created an equal mixture of good genes from cultivar X and not-so-good genes from cultivar Y (with the exception of the desired genes for fungal resistance). Backcrossing (see Figure 4.3) is a common approach to restore valuable characteristics. Several rounds of backcrossing are usually necessary before obtaining a plant that has all of the characteristics of the parent variety, but also the additional desired characteristic. This requirement for backcrossing is a major reason for the protracted length of time (and amount of labor) required for traditional breeding programs. Despite these problems, traditional plant breeding programs have been dramatically successful throughout the world. Virtually all of the food that we eat is from plants and animals that have been bred in this way, and newer techniques, such as transgenic development, likely will not substantially decrease the need for traditional breeding programs in the future. 2. Protoplast Fusion The major disadvantage of traditional breeding programs is that the range of sources of new characteristics is narrow; only closely related plants (usually within the same species) can be used. This is frustrating for plant breeders, who often see highly desirable characteristics in unrelated plants. Protoplast fusion partially solves this problem by allowing the recombination of chromosomes from a wider range of parents into a new cell, and ultimately, into a new plant. Protoplast fusion (Figure 4.4) often begins with mesophyll cells (mesophyll is the main photosynthetic tissue in leaves). The cells are treated with enzymes that break down the cell wall, resulting in protoplasts, plant cells that have an intact membrane but no cell wall. These cells can then be persuaded, through electroporation or the addition of polyethylene glycol (PEG), to fuse with protoplasts derived from other species of plants. When treated in this way, cells that collide undergo cell fusion. The membranes of two colliding cells fuse, and the contents of the cells are mixed. Nuclear fusion also occurs, and, if the plant breeder is fortunate, the cell will be able to cope with the mixture of chromosomes and divide mitotically. This usually occurs via the loss of chromosomes, but the resulting cell usually has some of the characteristics of the two original cells. Fused protoplasts can then be induced to re-form their cell walls and are then plated on agar media. If all goes well, fused cells will grow into a callus. Manipulation of hormones then leads to the formation of embryos (these are referred to as somatic embryos) that grow into plantlets. The resulting plants can then be backcrossed to restore valuable characteristics that were lost during the cell fusion process. Protoplast fusion has been less successful and less widely practiced than originally envisioned, primarily because of the difficulties in obtaining fused cells that will divide normally. However, this approach is useful if a desirable characteristic

90

Introduction to Food Biotechnology

Species A Digest cell walls Species B

Protoplasts PEG

Nuclear fusion

Cell fusion

Wall regeneration

Plant regeneration

FIGURE 4.4 Protoplast fusion between two sexually incompatible plant species. (PEG = Polyethylene glycol.)

resides in a species that is too distantly related to be crossed using conventional techniques, but is related closely enough to allow chromosomal re-assortment and successful division of cells fused with the parent variety. In plant breeding terms, then, protoplast fusion sometimes allows the use of wide crosses. 3. Somaclonal Variation: Problem or Opportunity? The manipulation of plant cells from plant to callus to plantlet was described as a relatively straightforward process. For some plants (e.g., tobacco) it is straightforward, but, unfortunately, the tissue culture of many plant species poses problems.

Plant Biotechnology

91

One of the problems shared by virtually all tissue culture systems is somaclonal variation. This is genetic variation that arises during callus culture. For reasons that are largely unknown (but much speculated upon), such variation frequently arises spontaneously, especially if callus culture occurs for a long time. In some cases, somaclonal variation can be useful. Plant breeders sometimes are frustrated by lack of variation. If there are no sources of a desired characteristic, then the breeder has nothing to start with. The parent crop, however, can be grown as a callus, broken up into individual cells, and diluted out onto an agar medium. Each of the resulting calli will then (if the dilution was successful) grow out from one original cell, just as bacterial colonies on a streaked plate grow from single cells. These calli can then be assessed for the presence of new, desirable characteristics. Although not widely practiced, this approach has resulted in the selection of improved plants (in one case, a crunchier carrot was obtained). Somaclonal variation is particularly useful if the desired characteristic can be detected during the callus stage. Otherwise, plantlets must be induced and transferred to soil before screening for the desired characteristic can proceed. Plant tissue culture is a complex art that has applications that are useful to food biotechnologists. However, there are many associated problems. Many important plants are difficult or impossible to grow in tissue culture. Furthermore, tissue culture is expensive in terms of materials and equipment, especially when compared to soilbased methods of plant culture. Despite these limitations, plant tissue culture is an important tool for plant biotechnologists. This is especially true for people interested in developing transgenic plants; tissue culture techniques are usually required at one or more stages in the development of transgenic plants.

III. TRANSGENIC PLANTS A. OVERVIEW Few subjects are as dear to the heart of a food biotechnologist (or as controversial to the public) as transgenic plants. These are plants that have received specific, welldefined genetic modifications. How is this different from traditional plant breeding? In a traditional breeding program, the changes that occur are usually undefined at the molecular level; the resultant plant might taste better, grow faster, and be more drought tolerant, but the breeder usually does not know exactly which genes have been gained or lost. In a transgenic approach, specific genes or groups of genes are transferred from a source organism into a target plant. Therefore, transgenic plants are conceptually similar to bacteria or fungi that have been given new genes using the techniques described in Chapter 3. Indeed, the techniques used in the development of transgenic plants are similar to the techniques used for the transfer of genes to bacteria and fungi. The major advantage of the transgenic approach over traditional approaches is that theoretically any organism can be a source of transferred genetic material. Genes can be transferred from distantly related plants, from bacteria, fungi, or viruses, and even from animals. Furthermore, the potential exists for exquisite control over the activity of transferred genes, in terms of the amount and the timing of gene expression.

92

Introduction to Food Biotechnology

This control is achieved through the use of specific promoter elements. Some promoters will allow expression only in certain tissues or organs (e.g., leaves, seeds, pollen, fruits) or in response to specific stimuli (e.g., wounding of the plant). This level of control can be useful. For example, consider a gene that codes for a protein that triggers the synthesis by plant cells of compounds that are toxic to fungi. Rather than have this gene continually expressed, it would be preferable to have expression only after wounding of the plant, and only in wounded tissues. Such tissues are vulnerable to attack by fungal pathogens, so the production of antifungal compounds would be beneficial. In contrast, the production of antifungal compounds in unwounded plants would be unnecessary and would waste energy and carbon. It might also be desirable to limit gene expression to parts of the plant that are not eaten, if there are any human toxicity concerns related to the antifungal compound. Most transgenic plants are developed using Agrobacterium tumefaciens, an unusual bacterial plant pathogen that has evolved the ability to transfer some of its genes to plant cells, where they are inserted into chromosomes to become a permanent part of the genome. Molecular biologists have modified this process to allow transfer of specific genes via A. tumefaciens, without the transfer of any genes of A. tumefaciens. Bacteria are the source of many of the genes that have been transferred to plants using A. tumefaciens. Bacterial genes conferring herbicide resistance and genes from Bacillus thuringiensis (Bt) have been transferred to numerous plants. In contrast, some transgenic plants are created through the insertion of genes native to the plant that have been altered in some way. These may be antisense versions of genes, which act to decrease expression of specific genes. For example, this is the usual approach taken to delay ripening in tomatoes and other fruits. Antisense genes decrease the levels of enzymes involved in ripening, such as enzymes that lead to ethylene production, and degradative enzymes such as polygalacturonase (PG), which lead to fruit softening. In some cases, the gene that is transferred originates from another plant. Transgenic canola varieties with high levels of laurate are an example. Laurate is a fatty acid that has many industrial (e.g., detergents) and food applications. Normally, canola oil does not contain this lipid; using A. tumefaciens, biotechnologists introduced a thioesterase from the California bay laurel tree into the canola genome, resulting in high levels of laurate in canola seeds. Most transgenic plants that have been released have been aimed at producers, especially toward the goal of easier and cheaper control of insect or weed pests. Biotechnology companies have concentrated on these applications because of the large, receptive market (agrochemicals are a major cost for most farmers, so they tend to be attracted to technology that allows decreased reliance on chemical control of pests) and because herbicide and insect resistance can often be induced through the addition of single genes. These types of transgenics will be more fully explored in Sections IV.A and C of this chapter, but several comments are appropriate here. Herbicides harm plants through disruption of crucial metabolic pathways (e.g., glyphosate, which disrupts synthesis of aromatic amino acids) or through a breakdown in hormonal control of plant growth (e.g., 2,4-D, which mimics the action of auxin hormones and results in uncontrolled growth that kills the plant). There are several ways to induce resistance to herbicides, but arguably the most common

Plant Biotechnology

93

approach is to find a bacterium that is able to degrade the herbicide, and then clone a gene for the enzyme that is responsible for this degradation. Then, transfer this gene to the plant using A. tumefaciens or an alternative method. The transgenic plant will then have the ability to degrade the herbicide before it can exert its toxic effects. This makes weed control easier and more efficient; fields can be sprayed when and if weed control is necessary, without having to worry about adverse effects on the crop. Insect control can also be given to plants through the addition of a single gene. There is a proviso, though: the specific insect pest must be affected by a strain of Bt. This soil bacterium produces proteins that are toxic to specific groups of insects; organic farmers have long exploited this property by spraying Bt spores onto crops. The transgenic approach is to clone the gene for a toxic protein from Bt and then to transfer the gene into the crop genome. Vegetative tissue of the crop will then be toxic to the insect pest, resulting in reduced use of chemical insecticides and fewer agrochemical costs to the farmer. At first glance, it may seem that herbicide- and insect-resistant transgenics are relevant to agriculture but not to food science and nutrition. However, the wide acreage of insect- or herbicide-resistant corn and soybeans, two commodities that are widely used in the food industry, makes these transgenics quite important. For example, the use of Bt transgenics should reduce the amount of pesticide residues on fruits and vegetables, thereby increasing food safety. Insecticide levels are generally low and considered to be safe, but in North America (and likely globally as well) there is a sporadic incidence of high levels of pesticide residues on fruits and vegetables that are consumed with minimal processing. Any factor that decreases reliance on pesticides is likely to decrease the incidence of such contamination. In contrast, the use of herbicide-resistant crops may not decrease the amount of herbicides used and in some cases leads to increases. However, the herbicides that are targeted by transgenics are typically less toxic and more rapidly degraded in the environment than many other herbicides. Therefore, the use of these transgenics may improve environmental quality and decrease the amount of toxic herbicide contamination of produce. However, this is a highly contentious issue, particularly in Europe, where there is a perception that farmers are much less reliant on agrochemicals than North Americans. Many people in Britain, for example, fear that the use of herbicideresistant crops will result in zero tolerance for weeds. This could result in a dramatic loss of plant biodiversity in British fields, which could have serious ramifications on wildlife. It is difficult to assess this risk, because it is difficult to assess the current “weediness” of British agriculture, and it is difficult to assess the importance of weeds in agricultural fields to wildlife. However, at least one study has shown that use of herbicide-resistant crops could theoretically reduce food availability (weed seeds) to skylarks. Similarly, the potential of Bt transgenics to reduce the use of chemical insecticides is controversial. With Bt corn that is toxic to the European corn borer, nontarget effects on monarch butterflies are an additional source of contention. Many scientists and antibiotechnology activists fear that pollen released by Bt corn will adversely

94

Introduction to Food Biotechnology

affect monarch larvae. This issue will be further explored in Section IV.A of this chapter. Consumers throughout the world are also concerned about safety issues associated with transgenic plants. Concern is particularly strong in Europe, where fear of bovine spongiform encephalopathy (BSE) and distrust of food regulatory agencies lingers. In 1998, mandatory labeling legislation was passed in Britain, making it illegal to sell food containing transgenic organisms (popularly known as genetically modified organisms, or GMOs) without labels. This led to the virtual disappearance of such food from British supermarkets, primarily because of fears by grocery chains that consumers would shun foods carrying GMO labels. Consumer groups and antibiotechnology activists are also pressing for labeling of meats and dairy products produced by cows fed feed containing transgenic crops. Many farmers in the U.K. are against such labeling requirements because of anticipated difficulty in obtaining sufficient transgenic-free feed. The situation in Europe and Britain illustrates the need for food companies and anyone that counsels people about food consumption to understand the techniques used to create transgenic crops, as well as the environmental safety issues that surround them. For example, a clinical dietitian may be interviewing someone who has recently been found to have high levels of cholesterol. The dietitian tells the client that s/he should increase consumption of fruits and vegetables. However, the client responds that he does not want to do that, because of “all those GMOs out there.” To respond effectively, the dietitian needs to understand how transgenic plants are made, as well as the risks and benefits associated with transgenic crops. Although most of the currently released transgenic plants have been aimed at producers, this will likely change in the next 10 years. Transgenics will be increasingly targeted to solve nutrition-related problems such as low vitamin content, high levels of antinutrients, amino acid imbalances, and allergenicity. The food industry will also benefit from transgenics with improved processing or storage qualities (see Chapter 1, Table 1.3). For example, the glutenin alleles present in a wheat variety strongly affect bread-making qualities of the variety. Transgenic technology allows fine control over the expression of these genes. These and other applications of transgenic technology to processing and nutritional aspects of food will be further discussed in Section V of this chapter. The main objective of the remainder of this chapter is to explain the techniques required for the development of transgenic plants and the range of opportunities associated with this technology.

B. THE PROCESS

OF

MAKING TRANSGENIC PLANTS

Consider the following scenario: you work for a biotechnology company that specializes in the production of transgenic food plants. You have been told that the company would like to consider developing an oilseed plant with high levels of a specific polyunsaturated fatty acid (PUFAx). What would be the overall plan of action? Although many variables should be considered, we can divide the process into pretransformation and posttransformation steps. In some cases, the posttransformation process may be more time consuming than the pretransformation process.

Plant Biotechnology

Pretransformation 1. Identify the target plant. Soybeans and canola are popular oilseed crops, as well as olive, corn, palm, and several other plants. Canola (low-erucic-acid rapeseed) is a popular choice for transgenic oilseed production because of the high oil content (40%) of the seeds and good agronomic qualities (e.g., cultivability over a wide geographic range). Canola is also relatively easy to grow in tissue culture and is easily transformed. 2. Identify the source plant. You need to find a plant that produces high levels of PUFAx. In many cases it will not be feasible to use this plant directly as a source of the oil, perhaps because it is not consumed as food, and thus has uncertain safety or edibility. However, it may be possible to transfer the trait to canola or another oilseed producer. 3. Isolate the enzyme that is responsible for the production of PUFAx. This will allow characterization of the enzyme’s properties and sequencing of the protein. Nucleic acid probes (Chapter 2, Section V.H.3) can then be designed, based on the amino acid sequence of the protein. 4. Isolate mRNA from the source plant and establish a cDNA library (Chapter 3, Section V). You will then be able to use the probe developed in step 3 to screen the cDNA library for the PUFAx gene. This will allow cloning of the PUFAx gene. 5. Confirm that the cloned gene is the correct gene by analyzing proteins expressed by the gene. This is normally done in a bacterial or fungal host cell. 6. Isolate genomic DNA from the source plant and set up a DNA library (Chapter 3, Section IV). Use the probe to screen this library for the desired gene. This is done so that the promoter and other genetic elements can be characterized. This is important, because transgenic oilseeds usually use promoters that restrict gene expression to the seed. Expression of fatty acid synthetic genes in other organs and tissues usually negatively affects plant growth. The promoter cannot be cloned from the cDNA library because the promoter region is not transcribed into mRNA. Sometimes, the researcher will already know of a promoter that will restrict gene expression to the appropriate tissue or organ. If so, this step may be less crucial. In most cases, though, it is advisable to know as much as possible about the gene and how it is regulated before proceeding to the next step. 7. Construct a plasmid vector (probably in Escherichia coli) that has the correct gene (from the cDNA library) and promoter (from the DNA library). 8. Insert this vector into the target genome, using A. tumefaciens, bolistics, electroporation, or micro-injection. These procedures virtually always require in vitro culture of the target plant. Following successful transformation, you will need to grow intact transgenic plants from these cultures.

95

96

Introduction to Food Biotechnology

Posttransformation 9. Reject transgenic plants that are clearly unsuitable, because of stunted growth or other abnormalities. Current methods of creating transgenic plants involve insertion of foreign genes into random sites within the plant cell’s chromosomes. This sometimes disrupts important genes, resulting in abnormal plants. The transgenic events that survive this screening process can then be assessed in subsequent steps. 10. Confirm that the desired gene is present in the transgenic plants. This can be done by southern blotting (Chapter 2, Section V.H.3). At this stage it is useful to determine how many copies of the gene have been inserted into the plant’s genome. In most cases, it is most desirable to have single-gene insertions, because this should lead to greater stability (see step 12). 11. Assess the phenotype of the transgenic lines. In other words, find out if the desired trait has been introduced. In our example, this would require analysis of the lipid composition of the seeds, to determine if PUFAx is produced. 12. Suitable lines are then tested for stability of gene inheritance and expression. This involves assessment of gene expression after crossing a transgenic line with nontransformed lines. It is essential that the recombinant gene be retained by at least some of the offspring of these crosses. 13. Use traditional breeding techniques to develop true breeding lines, so that all offspring of a transgenic plant will have the recombinant trait. This is impossible in some crops, but in such cases, alternative approaches such as hybrid crops are possible. Hybrid seed is made by crossing two varieties and selling the resulting seed (F1 generation) to farmers. Usually valuable traits are lost in subsequent generations, so every year the seed grower must repeat the cross, and every year the farmer must buy a new lot of seed. 14. Ensure that the resulting transgenic line has the necessary agronomic qualities (e.g., is able to grow in a specific range of soil types and climates) and processing characteristics (e.g., release of the oil through seed crushing). 15. Assess the environmental and food safety risks of the transgene (see Chapter 9 for a discussion of this process). This is required for the regulatory process in most countries. This should begin early in the process. It would be foolhardy to go to all the trouble and expense required to attain this stage without careful considerations of human and environmental safety issues. 16. Register the new transgenic variety (“line” and “variety” are similar terms, but variety is usually used at a later stage of the process). This process is similar to that undergone by varieties derived through traditional plant breeding methods. Registration normally involves demonstration to regulators that the variety is different from other varieties and that this difference is potentially useful to producers or processors.

Plant Biotechnology

97

Registration confers on the breeder certain intellectual property rights, but companies that develop transgenic plants usually attempt to gain patent protection as well. This gives them greater control over the use and sale of their transgenic varieties. 17. Obtain regulatory approval for release of the transgenic cultivar (a “cultivar” is a variety that is being actively cultivated). This normally involves submission of a document that assesses the characteristics of the transgenic protein and its potential toxicity and allergenicity to humans, as well as an assessment of potential environmental risks. In our example (PUFAx), the environmental assessment would likely be straightforward, but assessment of toxicity and allergenicity may be more difficult (see Chapter 9). In some regions (e.g., Europe), the company would have to obtain approval from two regulatory paths; one is set up to allow release of transgenic propagules (usually seeds) to farmers, and one is to allow use of the transgenic in food. 18. In many cases, it may also be necessary to ensure that the current path of the crop from the farm gate to processors can cope with the proposed transgenic variety. In our example, it would be necessary to segregate transgenic harvests from harvests of other cultivars (one would not want to mix PUFAx seeds with non-PUFAx seeds because it would result in dilution of the new valuable lipids). Segregation is often desirable for exporting purposes, as well; for example, shipments to Europe that are destined for use as food must be segregated into transgenic (GMO) and nontransgenic (non-GMO) lots, so that foods can be appropriately labeled. Also, most of the transgenic crops that have been approved in the U.S. and Canada have not been approved in Europe; in these cases, they must be segregated from European shipments, because it is illegal to sell them in the European Union. The development process for a transgenic plant is not fast, although in many cases it may be quicker than an equivalent process using traditional plant breeding, because of the precision of the genetic transfer (i.e., only specific genes are transferred using a transgenic approach). We will now concentrate on step 8, and develop an understanding of the methods used to transform plant cells with foreign genes.

C. AGROBACTERIUM

TUMEFACIENS:

A NATURAL DNA VECTOR

We now have a problem. Assuming that we have cloned our PUFAx gene into E. coli, how do we get it from the bacterial cell, through the plant cell wall, into the nucleus, and permanently integrated into the plant’s genome? Theoretically, this could be done in two ways: (1) by integration of PUFAx into a plasmid that is able to replicate in plant nuclei, or (2) by integration of PUFAx into a plant chromosome, which would allow PUFAx to be transferred into daughter cells through the normal segregation of chromosomes that occurs during mitosis. The second option is feasible for plant cells; unfortunately, plant cells rarely have plasmids, so the first option is not useful.

98

Introduction to Food Biotechnology

Plant is wounded

A. tumefaciens enters the wound and transforms cells

transformed cells grow to form a gall

A. tumefaciens is attracted to the wounded plant

FIGURE 4.5 Production of a gall by A. tumefaciens.

Getting foreign sequences of DNA into plant cells is not easy; obtaining a stable integration of foreign DNA into plant chromosomes is even more difficult. However, plant biotechnologists have been able to harness the ability of A. tumefaciens to act as a vector. This bacterium has been used to develop many of the transgenic plants that have been released to growers.

D. THE NATURAL LIFE CYCLE

OF

AGROBACTERIUM

TUMEFACIENS

A. tumefaciens is a Gram-negative aerobic bacterium that, until the 1970s, was known primarily as an unusual plant pathogen that induces tumor formation in plants. When it was discovered that it is able to transfer bacterial DNA into plant cells and plant chromosomes, plant biotechnologists and molecular biologists quickly saw the potential for harnessing the unique abilities of this bacterium. The life cycle of A. tumefaciens moves from a dormant to an active state when a plant is wounded (Figure 4.5). A. tumefaciens is able to survive free-living in soil, primarily in a dormant state. A number of compounds are released from plant cells after wounding, and A. tumefaciens is chemo-attracted to these compounds. It then enters the plant via the wound. Once inside the plant, A. tumefaciens induces tumor (gall) formation, usually in stem tissues. The bacteria reproduce extensively within the gall, and when the plant dies, bacteria are released into the soil, where the life cycle continues. A. tumefaciens is able to induce galls because of its large Ti (tumor inducing) plasmid (Figure 4.6). The vir and T-DNA regions of this plasmid are particularly interesting. vir contains genes that code for proteins that are able to direct and control the transfer of plasmid DNA from A. tumefaciens to plant chromosomes. However, the entire plasmid is not transferred; only the T-DNA is nicked out of the plasmid, transferred into the plant nucleus, and integrated into a chromosome. This is similar to conjugation, the process that allows bacteria to transfer plasmids from one bacterium to another. The transfer of DNA from bacterium to plant is complex and requires the coordinated activity of numerous genes in the vir region. Briefly, specific proteins recognize the DNA sequences of the left and right borders and excise a single strand of DNA from the region between the borders (the T-DNA region). This single-stranded DNA is coated with a DNA-binding protein and becomes part of the T-transporter complex (made up of 12 vir proteins), which moves the DNA from the bacterial cell into the plant cell. This apparatus probably moves through

Plant Biotechnology

99 Cytokinin synthesis

Auxin synthesis lb

Opine synthesis rb

T-DNA Opine utilization

vir

ori

FIGURE 4.6 Structure of the Ti plasmid of A. tumefaciens.

membrane channels out of the bacterial cell and into the plant cell. The complex moves from cell to cell via a pilus (T-pilus), a flexible, proteinaceous tube. Assembly of the T-pilus is mediated by vir proteins. Some of the proteins in the complex then function to allow transport of the complex into the nucleus. Once in the nucleus, the T-DNA is integrated (spliced) into a plant chromosome. The mechanism for this is unknown, but it probably involves host enzymes that normally function in DNA replication and repair. The transfer of T-DNA from bacterium to plant is an elegant, choreographed process requiring the participation of genes of the vir region, as well as several bacterial chromosomal genes and host plant proteins. Initially, scientists believed that the agrobacterium system of gene transfer is unique; however, it is becoming increasingly clear that many bacterial pathogens can transfer DNA and proteins into host plant, animal, or fungal cells. The T-DNA region of the Ti plasmid contains a number of genes that are necessary for the life cycle of A. tumefaciens. Several genes are involved in the synthesis of auxins and cytokinins, which trigger uncontrolled growth of plant cells (similar to callus formation in tissue culture systems). Other genes code for enzymes that synthesize opines, low-molecular-weight compounds that are an important source of carbon and energy for A. tumefaciens growing within galls. Opine production is a clever trick by the bacterium — each strain of A. tumefaciens produces a specific form of opine that can be catabolized only by that strain. Because most soil bacteria and fungi cannot utilize opines as a source of carbon and nitrogen, A. tumefaciens has exclusive use of these compounds. The production of a gall of plant cells producing opines, then, is in a sense a “private party” that only A. tumefaciens can attend.

100

Introduction to Food Biotechnology

MCS

Neo r

lb

rb

vir

ori A. tumefaciens ori E. Coli Vector 1

ori A. tumefaciens Vector 2

FIGURE 4.7 The binary vector system, derived from the Ti plasmid of A. tumefaciens. Most of the genes of the T-DNA have been removed, leaving the left and right borders. The vir region has been moved into another plasmid (vector 2). Note that plasmid 1 has an ori region that allows it to replicate in E. coli and an ori that allows it to replicate in A. tumefaciens (therefore, it is a shuttle vector). (MCS = Multiple cloning site, lb = left border, rb = right border, Neor = neomycin resistance.)

E. THE USE

OF

A.

TUMEFACIENS TO

CREATE TRANSGENIC PLANTS

It would clearly be unwise to use unmodified A. tumefaciens to develop transgenic plants. Such plants would have extensive galls and would not grow well. Consequently, many modifications have been made, particularly to the Ti plasmid. Most biotechnologists use a binary vector system (Figure 4.7), which uses two plasmids — one that contains the vir region (vector 2) and one that contains the left and right border regions of the T-DNA (vector 1). The major reason for constructing two smaller plasmids is that smaller plasmids are easier to manipulate; they replicate to higher copy numbers in bacteria, and it is much easier to get them into bacterial cells via transformation. Vector 1 contains the left and right borders of the T-DNA but none of the genes between these regions that are present in the Ti plasmid. All of the genes between the left and right borders of the T-DNA have been removed, so that tumors will not form. However, if a segment of foreign DNA is placed between the left and right borders, then the vir proteins will recognize the borders, excise the intervening DNA, and incorporate it into a plant chromosome. The following steps (Figure 4.8) are required for the transfer of a foreign gene (PUFAx, in our example) to a plant using an A. tumefaciens–derived binary vector system: 1. PUFAx is cloned into E. coli using an appropriate vector, as described in Chapter 3. The cloned gene is then subcloned into vector 1. Because the multiple cloning site (MCS) is between the left and right borders, PUFAx will be inserted between the borders. E. coli is used at this stage because it is easier to grow and manipulate than A. tumefaciens.

Plant Biotechnology PUFAx

101

lb

PUFAx vir

Neo r rb Vector 1

Vector 2

ori E. coli Clone in E. coli

Subclone in Vector 1

Induce plantlet formation

Transform A. tumefaciens with Vector 1

Infect callus tissue with recombinant A. tumefaciens

FIGURE 4.8 The use of the binary vector system of A. tumefaciens to transform plants.

2. Vector 1 is purified from E. coli and incorporated into A. tumefaciens via transformation. Plasmid 1 will be able to replicate in A. tumefaciens because it has an additional ori that allows replication in A. tumefaciens. The strain of A. tumefaciens used must contain vector 2, which has the vir genes required for transfer of PUFAx to plant cells. Such strains are readily available from a number of biotechnology companies. 3. A. tumefaciens now has vectors 1 and 2. The next step is to inoculate the bacteria onto cultures of callus derived from the target plant. A. tumefaciens will infect these cells, and plasmid 2 will produce proteins that will direct the transfer of PUFAx from plasmid 1 into plant cells. PUFAx will then be inserted into a plant chromosome. 4. The callus is transferred to a medium that contains neomycin (kanamycin). Plant cells are susceptible to neomycin, so any plant cells that lack the introduced foreign DNA will be killed. Vector 1 has a neomycinresistant gene between the left and right borders; if the transfer of PUFAx is successful, the kanamycin-resistance gene will also be transferred, and will be integrated into the plant genome. 5. The callus is then transferred to a medium that allows regeneration of plantlets. Many variations of the above scheme are possible. The inclusion of antibioticresistant genes in transgenics is controversial, so it may be preferable to use other marker genes. The GUS system is one alternative — this method requires that vector 1 contain a β-glucuronidase gene. Plant cells normally lack this gene, which codes for an enzyme that is able to convert X-glucuronide (a synthetic carbohydrate) to a blue end product. This color change allows identification of recombinant plant cells. Such cells can then be induced to grow into intact plants. GUS is called a reporter gene, because its presence indicates the successful transformation of a plant cell.

102

Introduction to Food Biotechnology

A. Bolistics DNA-coated particles

air gun

callus tissue

B. Electroporation

add desired DNA + PEG

dish with protoplasts

electric current leads to transformation of plant cells

C. Micro-injection

microsyringe with desired DNA nucleus protoplast

FIGURE 4.9 Alternatives to the use of A. tumefaciens to transform plant cells with foreign DNA. (PEG = Polyethylene glycol.)

Another common modification is to avoid callus culture and directly wound and infect plantlets. This avoids the genetic instability that often occurs during callus culture. This is particularly useful with plants that are difficult to manipulate in tissue culture. In some plants (e.g., rice), A. tumefaciens is able to infect only embryo cultures. These are cultures of cells that have been hormonally induced to form embryos.

F. ALTERNATIVES

TO

A.

TUMEFACIENS

The A. tumefaciens system works well with many dicots, but is less effective with monocots (e.g., cereals). For this reason, extensive research has been aimed at developing alternative systems for the transformation of monocots and other plants (Figure 4.9). One relatively successful system has been bolistics. By this method, DNA containing the desired foreign gene is attached to tiny beads, or microprojectiles. An air gun is then used to bombard plant cells or tissues with the microprojectiles. A small percentage of cells are successfully transformed in this way. Plants can also be transformed by electroporation of protoplasts and adding DNA after pretreatment of protoplasts with PEG. Finally, micro-injection of DNA into intact plant cells has been successful in some cases. One of the major disadvantages of all of these systems is that integration of foreign DNA into the plant genome rarely occurs; integration is much more frequent when A. tumefaciens is

Plant Biotechnology

103

used as a vector. Another disadvantage is that foreign DNA is inserted at random points in plant chromosomes. Consequently, disruption of essential genes sometimes occurs.

IV. APPLICATIONS OF TRANSGENIC PLANTS TO FOOD PRODUCTION A. TRANSGENICS RESISTANT

TO INSECT

PESTS

Many of the early transgenic crops were aimed at reducing reliance on chemical insecticides through the introduction of genes coding for insect-toxic proteins. This would have obvious benefits — farmers would be able to save large amounts of money through reduced purchase of pesticides, and the public would benefit through reduced consumption of pesticides in crops and decreased levels of pesticide residues in soil and water. The selection of a suitable gene for such transgenics is not easy. Three key criteria must be met: (1) the gene must code for a protein that is produced in plant tissues at a high enough concentration to be toxic to insect pests; (2) the protein must be toxic to insects but not to humans or other mammals; and (3) the protein should be highly specific, only killing targeted pest insects rather than a wide range of insects, because many insects are beneficial. Many proteins meet these criteria but proteins produced by the spore-forming bacterium Bt were the first to be exploited using transgenic technology. One reason for this is that Bt has a long history of safe and effective use in agriculture and forestry. As mentioned before, organic farmers rely heavily on Bt pesticides, applied to crops in the form of spore suspensions. These spores are toxic to insects because they contain crystalline proteins (both the proteins and the genes responsible are designated by cry). When certain species of insects consume the bacteria, the crystals become soluble in the alkaline environment of the insect midgut. After dissolving, they are converted from a protoxin to a toxin through the action of insect proteases in the midgut (the active toxin is designated by δ-endotoxin). cry proteins then interact with specific receptors on the surface of gut epithelial cells, leading to the formation of pores in the membrane. The epithelial cells then die, because of cytoplasm leakage. Extensive loss of gut epithelia eventually kills the insect. There are at least 135 different cry genes; they are classified into three groups (cryI, cryII, and cryIII), based partly on the size of the polypeptides and partly on the effective host range. The bacteria are also classified into groups (pathotypes) based on the effective host range of their toxins. For example, B. thuringiensis var. berliner produces cryI toxins, which are effective against Lepidopterans (butterflies and moths); B. thuringiensis var. israelensis produces cryII toxins effective against Lepidopterans and Dipterans (flies and mosquitoes); and B. thuringiensis var. tenebrionis produces cryIII toxins effective against Coleopterans (beetles). Fortunately, a number of important insect pests are Lepidopteran (e.g., the European corn borer), Dipteran (e.g., houseflies and mosquitoes), or Coleopteran (e.g., the Colorado potato beetle) and can potentially be controlled either with insecticidal sprays of Bt spores or through integration of cry genes into crop genomes.

104

Introduction to Food Biotechnology

δ-Endotoxins do not exert toxic effects on mammalian digestive systems. The safety of Bt toxins for humans is well established; large-scale spraying programs over large tracts of forest (aimed at control of spruce budworm) have shown no impact on humans, with the exception of uncommon allergic reactions to wholespore preparations. This is mainly considered to be a hazard to people spraying crops, and not to people consuming crops that have been sprayed with Bt spores. In comparison to chemical insecticides, Bt toxins are highly specific. Broadspectrum chemical insecticides kill beneficial predatory insects as well as insect pests. This often leads to problems due to secondary pests, insects that are normally controlled by predatory insects. Consequently, Bt insecticides are considered to be less damaging to the environment. The main factor limiting the implementation of Bt insecticides in conventional agricultural systems has been cost; chemical insecticides are generally cheaper. Biotechnologists quickly realized that this cost problem could be averted by creating transgenic crops that had cry genes integrated into their genomes. By the early 1980s, such transgenics had been successfully developed in the laboratory, and in 1996, the U.S. FDA allowed the release of transgenic insect-resistant corn carrying cry genes. In 2000, nearly one third of the American corn crop was Bt corn, a phenomenal success story, at least in terms of its rapid adoption by farmers. However, early estimates of the 2001 crop suggest that use of Bt corn fell slightly in 2001, possibly because of fears that major world markets (e.g., Europe) would not buy very much Bt corn.. Technically, the development of Bt transgenic plants is not easy. It is relatively easy to clone cry genes and transfer them to plants (both bolistics and A. tumefaciens have been used, depending on the target plant). However, early attempts resulted in plants with only a small amount of expression of the toxic protein — not enough to control the pest. Scientists soon realized that this low expression was partly due to codon usage. Remember that more than one base triplet (codon) can code for each amino acid. Unfortunately, bacteria tend to prefer certain codons and plants tend to prefer other codons for certain amino acids. Consequently, biotechnologists had to modify the DNA sequence of the cry genes to reflect this difference in codon usage. This, combined with the use of effective promoters, led to large increases in the amount of δ-endotoxin in plant leaves. The promoter from cauliflower mosaic virus (CaMV) has been used in some Bt transgenics, leading to high rates of expression in all tissues of the plant. Other transgenics have used promoters that restrict expression to either leaves or pollen. The introduction of enhancer elements (see Chapter 2, Section V.C) has also been effective in increasing levels of δ-endotoxin expression. Transgenic Bt crops have been successful from a technical perspective (i.e., they effectively control target pests) and from a commercial perspective. However, Bt transgenic plants (especially Bt corn and Bt cotton) have also been attacked by a number of environmental and antibiotechnology activist groups. One controversy centers on the potential of target insects to quickly develop resistance to B. thuringiensis toxins. Organic farmers would lose one of their most valuable weapons against insect pests (organic farmers usually restrict their use of Bt sprays, partly for economic reasons and partly to reduce the risk of encountering resistant insects). The potential to develop resistance is a concern because: (1) resistance to B. thuringiensis

Plant Biotechnology

105

toxins has already been observed in several instances, following long-term, intensive spraying programs, and (2) large-scale growth of Bt transgenics creates an environment that should efficiently select resistant insects. The response of the biotechnology industry to these criticisms is that resistance may occur, but it can be countered by the introduction of new transgenics, using different toxin genes, or stacked transgenics that contain more than one cry gene. Also, companies and regulators currently rely on the high-dose/refugia approach to delay resistance development. Refugia are areas with non-Bt plants within or adjacent to Bt plantings. Theoretically, the use of high-dose insecticides combined with refugia should delay resistance development. The reasons for this (and the use of refuges) are complex and involve the heritability of resistance genes. Recall that classical mendelian genetics describes genes as either dominant or recessive. If resistance genes are recessive, then refugia will allow Bt-sensitive (ss) insects to persist. They should breed with resistant (SS) insects in the Bt part of the field, leading to heterozygous (Ss) offspring. If the phenotype (resistance to Bt) is recessive, Ss insects should be susceptible to Bt. Refugia act as reservoirs of sensitive insects. However, if resistance is dominant, then Ss insects will be resistant to Bt. This is where the high dose becomes important (Figure 4.10). If the level of Bt in plant tissues is very high, then Ss insects will be susceptible, even if resistance is a dominant trait. This is an important point, because it is currently impossible to predict whether insect resistance will be dominant or recessive; both have been observed after intensive spraying of Bt spores. This illustrates the practical importance of the development of Bt transgenics that have high levels of expression of the Bt toxin. In the next 5 years or so it should be possible to assess the success of the highdose/refugia strategy. As an example of refugia, in 2000, the EPA recommended that U.S. corn growers establish refugia constituting 20% of their Bt corn fields and that they monitor their Bt corn closely, to establish an early warning system to detect resistant populations. Resistance has appeared in some regions; for example, surveys of resistance to Bt cotton in populations of cotton bollworm in Australia revealed that resistant insects are present and are slowly increasing in number. The slow increase may be partly due to adverse aspects of resistance on the insects (e.g., reduced number of eggs produced) that weigh against the favorable aspects of resistance (i.e., ability to feed on Bt cotton). Because of insect resistance, each Bt cultivar may have a relatively short life span in agricultural use. However, if resistant insects are less fit than nonresistant insects, it is possible that temporary removal of a particular Bt cultivar could result in a loss of resistant insects. If this happens, the cultivar could then be reintroduced. It should be stressed, though, that resistance to δ-endotoxins is relatively poorly understood, making it difficult to predict when and where resistance will occur, and how stable it will be. The second controversy involving Bt corn concerns its effects on nontarget insects, particularly spectacular and popular insects such as monarch and swallowtail butterflies. These insects have several reproductive cycles every summer in the American midwest, leading to the occurrence of larvae feeding in and around corn fields. They do not feed on corn leaves, but they may accidentally ingest corn pollen

106

Introduction to Food Biotechnology

Non-Bt crop is planted in adjacent refugia

Bt transgenic crop is planted in a field

Insects are sensitive to Bt (ss genotype)

A small number of insects develop resistance (SS genotype) Both populations interbreed

ss genotypes have an advantage over Ss (resistance carries a cost)

Few SS genotypes are present. Most insects are ss, and a small number are Ss

Few Ss individuals survive. Most of the population is ss

Ss individuals are killed by the high dose of Bt

Most of the population is ss

FIGURE 4.10 The theory behind the use of refugia and high doses of Bt to discourage the spread of resistant insects. This system should work even if resistance is a dominant trait in insects, as shown here.

containing δ-endotoxin by eating leaves of milkweed plants (main food of the larvae) that have corn pollen on their surfaces. Indeed, several laboratory studies have demonstrated toxic effects of Bt corn pollen when fed to monarch butterfly larvae. The investigators placed milkweed plants within and on the edge of Bt corn plots during the period of pollen release. Then they fed samples of the milkweed leaves to monarch larvae reared in the laboratory. Twenty percent of the larvae subsequently died, whereas none of the insects died that were fed leaves exposed to non-Bt corn pollen. This raised a number of questions regarding the environmental safety of Bt corn because of the following points: • Monarch butterfly larvae feed on milkweeds, which are commonly found within and adjacent to corn fields. • Some Bt corn cultivars produce high levels of toxic proteins in the pollen; in some cases this is done deliberately through the use of pollen-specific promoters, because the European corn borer tends to consume large amounts of pollen.

Plant Biotechnology

107

The monarch butterfly is arguably the most popular insect on earth, and any threat to it raises the hackles of conservation groups. It has become the rallying cry of antibiotechnology activists throughout the world, and it is unlikely that hostility to Bt corn in Europe and elsewhere will abate without clear evidence that Bt corn does not harm the monarch butterfly. Biotechnologists have not been silent on this issue; many have argued that the evidence against Bt corn is inconclusive, for the following reasons: • To date, no field study has demonstrated harm to monarch butterflies from ingestion of pollen from Bt corn. The environment in the field is much more complex than the controlled environment of the laboratory, and scientists often observe that effects observed in the laboratory do not occur in the field. In the case of the monarch butterfly, factors such as predation and lack of milkweed plants may be much more important than toxicity of Bt pollen. • Large amounts of pollen are required to exert toxic effects on larvae; such large amounts are unlikely to occur anywhere except the interior of Bt corn fields, where the density of milkweeds is usually low. • Toxicity to larvae is highest in transgenic cultivars with pollen-specific promoters; these tend to be less popular than transgenics with less specific (e.g., CaMV) promoters. • The alternative to using Bt corn is to use chemical insecticides to control the European corn borer; these insecticides kill all lepidopteran larvae, including the monarch butterfly. This is an area of active research, and it is anticipated that studies will soon be published that will help resolve this controversy. Until then, it is unwise to claim that Bt corn is either good or bad for monarch butterflies. Pro-Bt forces were recently encouraged by the publication of a study that demonstrated that Bt corn was not toxic to larvae of the swallowtail butterfly, another beautiful and popular butterfly in North America. The third controversy is widely known as the Starlink scandal. In October 1999, it became clear that food in the U.S. had become contaminated with corn from a Bt cultivar (Starlink corn) that the FDA had allowed to be used as animal feed but not as food for humans. The FDA had made this decision based on the potential allergenicity of this corn, because the insect-toxic protein was less degradable under simulated stomach conditions (this is a characteristic of many food allergens). The main company involved (Kraft) lost millions of dollars because of a recall of contaminated products such as taco shells. Trade to Japan and other countries has also been disrupted after the detection of Starlink corn in exported corn. The Starlink scandal, and its ramifications, will be more fully explored in Chapter 9.

B. PATHOGEN RESISTANCE Plants suffer from many diseases; some are physiological, caused by drought stress, mineral deprivation, and other environmental causes, but infectious agents (pathogens) also cause disease, reducing the amount of harvestable food by about 15%

108

Introduction to Food Biotechnology

globally. Post-harvest disease (spoilage) is also significant, resulting in the loss of about 25% of harvested food. In declining order of importance, fungi, viruses, and bacteria are the pathogens responsible for infectious plant disease, and fungi are responsible for most post-harvest food spoilage. Besides the economic losses, field and post-harvest diseases are a safety issue for the food industry and for consumers. Many of the fungi commonly involved produce mycotoxins. For example, Aspergillus flavus, which produces potent mutagens known as aflatoxins, is commonly found on cereal grains and peanuts grown in tropical or subtropical countries. Most countries have defined limits of aflatoxin contamination permitted in food and feed. For example, the U.S. has set limits of aflatoxin contamination in nuts destined for human consumption at 20 µg/kg. Much of the fungal contamination that leads to mycotoxin contamination or food spoilage arises in the field; therefore, efforts to control fungal attack in the field may lead to improved food quality after storage. Interestingly, this is one of the points raised by proponents of Bt transgenic corn — when the European stem borer attacks corn plants, it creates a route of entry for fungi. It has been shown that levels of mycotoxins are lower on grains of Bt corn than on nontransgenic corn. Other transgenic crops are being developed that are better able to resist the growth of mycotoxigenic (capable of mycotoxin production) fungi such as Aspergillus and Fusarium spp. (the latter is a fungus commonly found in cereals grown in temperate countries). Plant improvement is also the main strategy in the fight against plant disease in the field. The wheat stem rust epidemic that devastated wheat production in the mid1950s in North America is an example of both the destructive ability of field plant pathogens and the potential of plant breeding to decrease the damage. The fungus that causes wheat stem rust (Puccinia graminis) produces small spores that can be transported via wind currents over wide distances. This allowed stem rust to spread over wide regions of North America, causing great loss of wheat yield. A concerted effort by Canadian and U.S. plant breeders led to the development of wheat cultivars resistant to the major pathovars of P. graminis. Traditional plant breeders have successfully incorporated resistance genes into most crops. This resistance works through the incorporation into a plant of a gene that allows recognition of a fungus within plant tissues. The plant then reacts quickly (e.g., through the production of phyto-alexins that are toxic to the fungus) to eliminate the pathogen. It is possible to transfer resistance genes from plant to plant using transgenic technology. This is an attractive option for many plant–pathogen combinations, because resistance genes are often present in wild plants that normally cannot be bred with crop plants. With the advent of gene cloning, and interplant transfer using A. tumefaciens and other methods, resistance genes can now be transferred from one plant to any other plant. As an example of transgenic pathogen resistance, we will examine transgenic plants that are resistant to specific plant viruses. Viruses pose unique problems to farmers. They cannot be attacked directly using pesticides, and antiviral drugs are not available for use on plants. Furthermore, the antiviral drugs used for human and

Plant Biotechnology

109

animal therapy are prohibitively expensive, and therapy of individual plants is impractical, due to the huge numbers of individual plants within a field. This is an important problem, because of the economic losses suffered worldwide as a consequence of viral infection of crop plants. Current strategies in the fight against plant viruses are based on prevention. The most effective way to prevent costly virus infections is to use plant cultivars that are resistant to viral infection; unfortunately, resistant cultivars are not available for some crops. In these cases, the only alternative is to ensure that crop propagules (seed, in most cases) are virus free, and to control insects such as aphids that spread viruses from infected fields to noninfected fields. Consequently, current virus-control strategies rely heavily on chemical insecticides, with their associated risks, including occupational exposure of farmers to concentrated toxic solutions, insecticide residue on food, damage to nontarget insects, and accumulation of toxic residues in soil and water. For these reasons, there is much interest in the development of virus-resistant cultivars. Several transgenic cultivars belonging to the Cucurbitaceae family (e.g., squash, watermelon) have been released in the U.S. Transgenic virus resistance is usually based on transfer of genes coding for either viral coat proteins or viral replicase genes. Replicase genes work through the phenomenom of gene silencing whereby the addition of a specific viral gene to the plant genome results in loss of expression of that viral gene when the virus infects the plant. The mechanism behind this silencing is unclear. More commonly, viral coat protein genes are introduced into transgenic virusresistant plants. Why are these genes effective? For a virus to replicate, it must take over the host cell and direct the cell’s metabolic machinery toward the production of new viral particles (Chapter 2). Before they can exert this control, viruses that infect eukaryotic cells must first be uncoated, so that the viral genome can be released into the cytoplasm. This uncoating (see Figure 2.8) is essential, because viral control of the cell requires the expression of viral genes. Transcription of viral nucleic acids cannot occur unless they are liberated from their protective coat of capsid protein. Therefore, if uncoating is blocked, the virus will be unable to replicate. For reasons that are unclear, large amounts of coat protein in a plant cell inhibit uncoating, and therefore give a plant a high level of resistance against viral infection. This resistance is highly specific; for example, a transgenic tomato that expresses the coat protein gene of tomato mosaic virus (ToMV) will be resistant against ToMV, but not any other virus. Therefore, if a particular crop is affected by a number of viruses, complete virus protection requires the addition of a large number of coat protein genes. Because this kind of protection requires high levels of expression of the transgene, the presence of numerous different coat protein genes may lead to unacceptably high levels of viral proteins within plant cells. The major controversy surrounding virus-resistant transgenic plants lies in the presence of viral genes in crop plants. Certain scientists and public groups are concerned that such genes could mutate into more virulent forms and be incorporated into viruses, leading to more virulent forms of the virus. This has not occurred yet and does not appear likely.

110

Introduction to Food Biotechnology

C. HERBICIDE RESISTANCE Not surprisingly, the research leading to these transgenics has for the most part been funded by companies that manufacture herbicides. The driving force behind this research is the cross reaction that often occurs when herbicides are applied to growing crops — herbicides often damage crops as well as weeds. Because of this, farmers are forced to alter their herbicide application strategies. They may apply herbicides pre-emergence (before crop seedlings have germinated and emerged from the soil), or they may use herbicides that are less powerful but do less damage to crop plants. From the point of view of both farmers and herbicide companies, it would be useful to be able to grow plants that are resistant to herbicides; the herbicides can then be applied when the crop is actively growing, when weeds are most likely to interfere with crop growth. How are herbicide-resistant plants developed? A variety of strategies are possible, but the most popular approach has been to manipulate herbicide targets. Herbicides often inhibit specific enzymes; for example, the broad-spectrum (affects most plants) herbicide glyphosate inhibits 5-enolpyruvylshikimate phosphate (EPSP) synthase, an enzyme that plants require for the synthesis of aromatic amino acids. Glyphosate (known commercially as Roundup™) is structurally similar to the normal substrate (phosphoenol pyruvate) of this enzyme. This similarity allows glyphosate to act as a competitive inhibitor of EPSP synthase; the plant is then unable to synthesize aromatic amino acids, and it dies. There are several ways of producing transgenic plants that are resistant to glyphosate. One strategy is to add additional copies of the gene coding for EPSP synthase behind strong promoters; the level of EPSP synthase within plant cells may then become high enough to overcome the inhibition caused by the herbicide. Another successful approach has been to introduce genes that code for proteins that degrade specific herbicides. Such genes are often found in soil bacteria and fungi. Finally, some glyphosate-resistant transgenics have genes coding for bacterial versions of EPSP synthase that are unaffected by glyphosate. All three strategies have been successful. A small number of herbicide-resistant crops have also been developed through traditional plant breeding. Opponents of herbicide-resistant transgenics believe that the use of such transgenics will lead to increased herbicide use, resulting in increased contamination of food and the environment with pesticide residues. The rejoinder from the herbicide companies is that it is not their intention to increase herbicide use, but only to increase the market share of their company. They also maintain that herbicide use will not increase, but will become more efficient, and that herbicide-resistant transgenics allow farmers to use less toxic, more biodegradable herbicides such as glyphosate. In 2001 the Centre for Agriculture and Environment (CLM) published an analysis, based partly on data provided by the U.S. Department of Agriculture (USDA), of herbicide use following the introduction of herbicide-resistant soybeans to the U.S. Farmer experiences varied, but the CLM concluded that these transgenics result in a mild decrease (0 to 10%) of herbicide use. The possibility of creating “superweeds” has also received much attention, particularly in the popular media. Antibiotechnology activists fear that the cultivation

Plant Biotechnology

111

of herbicide-resistant transgenics will result in weeds that cannot be controlled which will cause crop loss and invasion of natural habitats. Regulatory agencies in the U.S. and Canada considered this risk when assessing transgenic herbicide-resistant soy and canola. In Canada, soy represents less of a risk than canola, because soy lacks weedy relatives that it could successfully pollinate (hybridize). Hybridization in this context refers to the formation of viable offspring from the union of gametes from a transgenic plant and a related species. In contrast, the type of canola used to develop transgenics (Brassica napus) can hybridize with seven related Canadian species, including Brassica rapa (the “other” canola), which is a weed in the Canadian prairies. The Canadian Food Inspection Agency (CFIA) allowed the release of transgenic herbicide-resistant canola, despite the potential for spread of the gene to other species, for two major reasons: (1) acquisition of the resistance gene by weedy relatives would only benefit these relatives if they were exposed to glyphosate, and (2) glyphosate is not normally used to control these weeds. In other words, farmers that do not grow herbicide-resistant crops would not normally use these herbicides, so the only risk would be to growers using the transgenics. Presumably, then, if glyphosate-resistant weeds become a nuisance, farmers can use methods of weed control appropriate for nonglyphosate-resistant cultivars. Increased weediness of the crop is another risk that must be assessed. Crops acting as weeds (volunteers) can sometimes be a problem. Weediness is controlled by a complex number of traits, including the time it takes for seed production, the number of seeds produced, the ability of the plant to grow vegetatively (e.g., by horizontal stems), and the ability of seeds to overwinter. It is unlikely that the transfer of herbicide-resistant genes will affect such developmental traits. However, regulators usually require documentation that some of the weed-related traits (e.g., seed production) are not different in the transgenic line. Plant biotechnologists were encouraged by a recent study of the overall weediness of a number of insect- and herbicide-resistant cultivars in 10-year-old experimental plots in the U.K. The transgenic crops had disappeared from all but one of the plots, allaying fears that transgenic plants would increase weed tendencies.

D. DELAYED RIPENING In 1994, Calgene received permission from the FDA to release the first transgenic food, the Flavr Savr™ tomato. Although this transgenic has not been very successful commercially, many biotechnology companies remain convinced that controlling ripening through transgenic technology is a viable idea. Fruit ripening is an essential but problematic process. Ripening is essential to food production because unripened fruit is usually inedible. Ripening is associated with changes in the texture, structure, and flavor of fruit. From the point of view of the consumer, ripened fruits are softer, taste much better, and are easier to digest than unripened fruit. These positive changes are also advantageous for the plant. Many plants depend on animals for seed dispersal — animals eat fruits and scatter seeds around, either because of their messy eating habits or through the safe passage of seeds through the animal’s digestive tract. The best way to encourage animals to

112

Introduction to Food Biotechnology

TABLE 4.1 Changes that May Occur during Fruit Ripening and Their Relevance to Food Quality Change

Relevance to Food

Loss of chlorophyll Pigment accumulation Change in cell wall structure Changes in organic acids Increased production of volatile chemicals

Less green appearance Fruit color development Softening Flavor changes Flavor changes

eat fruit is to make fruit tasty and nutritious. Evolution, therefore, has favored fruits that taste good and are soft and easy to chew and swallow. People have expanded this trend through the selective breeding of plants. This combination of soft and tasty fruit is great for the consumer, but it causes problems for food producers and processors. These problems are particularly acute in North America. Most of North America’s fruit is produced in Mexico, California, or Florida and transported by rail or truck to points throughout North America. Many bumps and shocks are inevitable with this transportation, and the stress is exacerbated when fruit is transported overseas, as is the case with tropical fruits such as mangoes and bananas. Because of this transport stress, fruit producers have two choices: they must grow fruit cultivars that produce hardy fruit that can withstand transport, or they must transport fruit in an unripened state, when it is much more resistant to physical shocks and stresses. The transport of unripened fruit is a popular strategy. Some fruits, such as oranges, ripen quite well after transport; others, such as tomatoes, ripen much more efficiently if ripening is induced by the application of the plant hormone ethylene. Unfortunately, most tropical fruits (bananas are an important exception) cannot be induced to ripen and cannot be transported in the ripe state, making it virtually impossible to transport these fruits to distant markets. Storage of fruits is also a problem. Some fruits (e.g., apples) can be stored for long periods in controlled atmospheres (e.g., reduced O2, increased CO2, 4°C). What happens during fruit ripening? In climacteric fruits, ripening is associated with a burst of respiratory activity triggered by increased production of ethylene in the fruit. In contrast, nonclimacteric fruits do not have a burst of respiration at the onset of ripening, and ethylene does not trigger or accelerate the ripening process. In both fruit types, ripening is associated with a number of biochemical and structural changes (Table 4.1) that strongly affect fruit quality and perishability. In relation to perishability, softening is the most important factor. It is caused principally by degradation of pectin by polygalacturonase (PG), with the help of a number of other hydrolytic enzymes. Pectin is a heterogeneous polysaccharide composed mainly of a backbone of 1,4-linked galacturonan, with side chains of rhamnose residues. It acts as a cement in which cellulose microfibrils are embedded, producing a rigid, strong wall. When pectin is broken down, this strength and rigidity dramatically decline, resulting in fruit softening (Figure 4.11). Individual cells within fruit

Plant Biotechnology

113

plant cell cell wall

middle lamella

FIGURE 4.11 Polygalacturonate in plant cell walls. The cell wall contains pectic acid, and the middle lamella is particularly rich in pectic acid.

Sense Gene 3’

Promoter TTCAGCATTCG

5’

3’

Antisense Gene Promoter CGAATGCTGAA

5’

transcription Sense mRNA

5’

3’

AAG UCGUAAGC UUC AGCAUUCG

3’

5’

5’ 3’ GCUUACGACUU Antisense RNA Flip the antisense RNA over to show the complementary association with the sense mRNA

FIGURE 4.12 The relationship between a sense gene and an antisense gene. An antisense gene is constructed by reversing the sequence of the gene. This results in a mRNA product that is complementary to the sense mRNA. Association between the sense and antisense mRNA results in decreased translation of the sense mRNA and lower levels of the protein in the cell.

also lose adhesiveness because the pectin-rich middle lamella between cells is responsible for cell–cell adhesion. This leads to further fruit softening. Transgenic fruits with delayed ripening have been made using two strategies: (1) reducing the amount of PG present in ripening fruits, and (2) reducing the amount of ethylene produced (ethylene triggers PG production in climacteric fruits). Antisense technology is usually the method of choice for both purposes. An antisense gene (Figure 4.12) is constructed by reversing the sequence of the gene. Transcription of an antisense gene produces an RNA molecule that is complementary to the sense mRNA sequence. Note from Figure 4.12 that the sense and antisense RNA strands are antiparallel and thus can form a double-stranded RNA molecule similar to double-stranded DNA.

114

Introduction to Food Biotechnology

The formation of this double-stranded RNA molecule leads to decreased translation of the sense mRNA. The reasons for this are not fully understood but may involve host cell defense systems against viruses. Double-stranded RNA is a red flag for cells, because this type of RNA is usually only present in virally infected cells. Therefore, it is not surprising that enzymes that recognize and degrade doublestranded RNA are common in eukaryotic cells. Antisense DNA may also work through gene silencing; under this scenario, DNA–RNA hybrids of a gene trigger methylation of that gene. Once a gene is methylated, it will not be expressed. Alternatively, antisense DNA may work through impaired translation of the endogenous (normal) gene. Formation of sense–antisense double-stranded RNA molecules may prevent translation by the ribosome. The Flavr Savr tomato has lower levels of PG and consequently a decreased rate of softening after harvest. However, the difference in softening appears to be modest, probably because of the role of other wall-softening enzymes. Inhibition of ethylene production is an alternative approach to inhibition of softening. Plants produce ethylene from a methionine precursor; one of the enzymes involved (1-aminocyclopropane-1-carboxylic acid [ACC] oxidase) has been targeted through antisense technology. Transgenic tomatoes with antisense ACC oxidase genes have much reduced rates of softening and do not noticeably ripen. Fortunately, these transgenics will ripen after the application of exogenous (from outside of the plant) ethylene, allowing transport of the plants in an unripened (preclimacteric) state. After transport, the tomatoes are exposed to ethylene and ripening is triggered. These tomatoes are unlikely to offer improved tomato flavor or color to consumers, but the main application is to reduce the amount of fruit loss that normally occurs during transport. In conventional tomatoes, a substantial proportion of preclimacteric fruits will begin to ripen during transport, resulting in damage. With transgenic tomatoes that produce low levels of ethylene, this should not happen. Tomato processors can also benefit from transgenic technology. Transgenic tomatoes with decreased levels of PG have increased Bostwick viscosity, a variable positively correlated with paste yield (the amount of paste obtained from a given number of tomatoes). Transgenic tomatoes with reduced levels of both PG and pectinesterase (another wall-softening enzyme) have increased levels of Bostwick viscosity, serum viscosity (this gives a more glossy paste), and soluble solids. All of these traits are valuable to tomato processors and may also be relevant to consumers. In the mid-1990s, one of the most popular brands of tomato paste in the U.K. was derived from transgenic tomatoes with decreased levels of PG. This tomato paste was perceived by consumers to be thicker and more flavorful than other pastes. Some chefs were also enthused by its ability to stick to pasta. Unfortunately, because of grocery chains’ fears of consumer reactions to transgenic foods, this brand is no longer available in the U.K. Because the Flavr Savr tomato was the first transgenic food, much attention focused on its safety assessment. It was the first transgenic crop to undergo an assessment of substantial equivalence (see Chapter 9 for a discussion of this concept). Under this strategy, we compare the levels of nutrients, toxins, and vitamins in the fruit to the parent variety. Calgene demonstrated that the only detectable differences between the transgenic tomato and the original tomato were in the levels

Plant Biotechnology

115

of polygalacturonase (lower in the transgenic) and the rate of softening (slower in the transgenic). In other respects the tomato was unchanged; it was nutritionally similar and had similar levels of toxins (tomatoes normally contain low levels of tomatine, a toxic alkaloid). This is perhaps an opportune point to discuss the presence of antibiotic-resistant genes, because this is the major criticism raised against the Flavr Savr tomato. As discussed previously in this chapter, antibiotic-resistant genes are included with the transgene so that selection of recombinant plants can be achieved simply by placing the plant cells on an agar medium containing the antibiotic. However, the use of antibiotic-resistant genes in this way leads to potential risks that need to be carefully assessed. Ingestion of plants containing antibiotic-resistant genes could theoretically lead to increased frequency of antibiotic resistance among bacteria in the gastrointestinal tract. This is certainly conceptually possible; DNA in plant cells could be released from lysed plant cells in the gut, and bacteria could then gain this DNA through transformation. Fortunately, successful transformation is unlikely. Relatively few bacteria are naturally competent (i.e., able to be transformed without treatment with electric currents or chemicals such as CaCl2). The human colon is dominated by enterococci that do not appear to be capable of natural transformation. Finally, even if a bacterium in the gut became transformed with an antibiotic-resistant gene, this bacterium is unlikely to cause problems unless the person is taking the specific antibiotic that the resistance gene is effective against. If so, then resistant bacteria would be heavily favored, as compared to bacteria lacking the resistance gene. Kanamycin, the aminoglycoside antibiotic used most commonly in the selection of successful transgenic plant cells, is not widely used for therapy of microbial infections in humans. For these reasons, the use of antibiotic-resistant genes is not generally viewed by the scientific community as a dangerous practice. However, because of public concerns, and because any dissemination of antibiotic-resistant genes is unwise, many biotechnologists are using or developing alternatives to antibiotic-resistant genes. The GUS system, which allows the detection of recombinant plant cells through a visual change in recombinant cells, is a popular alternative. Unfortunately, because this is not a selective process (i.e., it does not eliminate nonrecombinant cells), it is technically less efficient than the use of an antibiotic resistance strategy. There are several other alternative strategies that can eliminate antibiotic-resistant genes in transgenic plants. For example, sequences can be introduced around antibiotic-resistant genes that are recognized by enzymes involved in the excision of transposons from DNA. The antibiotic-resistant gene can be used in the normal way to select transformed plant cells, and the transposon machinery can then be used to remove the antibiotic-resistant gene. This and other strategies are increasingly being used in the development of transgenic crops.

E. GOLDEN RICE In 1994, the World Health Organization (WHO) estimated that 2.8 million children ages 0 to 4 suffer clinically from vitamin A deficiency disorder (VADD). Clinical deficiency is evident through several types of vision problems (xerophthalmias), such as night blindness and scars on the cornea. In severe cases, vitamin A deficiency

116

Introduction to Food Biotechnology

leads to blindness. Subclinical deficiency is also globally widespread — in such cases, children and adults (particularly pregnant or lactating women) have decreased levels of retinal (a form of vitamin A) in their serum. They may not exhibit symptoms of vitamin A deficiency, but they are at high risk to develop symptoms. The WHO estimated in 1994 that 251 million children are subclinically affected. The United Nations and others have been aware of this ongoing global tragedy since the 1960s, and concerted action has resulted in decreased incidence of clinical VADD, particularly in India, Bangladesh, and Indonesia. This decrease has been primarily driven by the widespread dispersal of vitamin A tablets. Unfortunately, VADD is still prevalent in many areas of the world and is most acute in subsaharan Africa and southeast Asia. VADD is difficult to solve globally because it can be caused by a number of different factors. However, people suffering from VADD often fall into one or more of the following categories: • Communities that use rice as the staple food. In many parts of the world, especially southeast Asia, rice is considered to be a satisfying and adequate food for young children. Unfortunately, milled rice has very low levels of β-carotene, one of the most important provitamin A compounds. Communities that use wheat as a food staple suffer lower levels of VADD than those using rice. • Low socioeconomic status. In several countries, there is a strong negative correlation between income and incidence of VADD. Poverty results in an overreliance on cheap staple foods such as rice. • Periodic food shortages. In drought seasons, or drought years, food shortages may lead to dramatically increased levels of VADD. For example, in the desert region of Rajasthan, India, a single drought year in 1987 led to an increase in xerophthalmia from 10 to 35% of the population. Humans and other vertebrate animals produce vitamin A mainly through the conversion of dietary sources of carotenoids. The most common carotenoid thus used is β-carotene. This yellow-orange plant pigment is converted by the human body into all-trans retinal (Figure 4.13), a fat-soluble vitamin that is required for vision because of its role in the development of the cornea and its function in transmission of electrical signals from the retina to the brain. 11-cis Retinal (a compound derived from retinol) binds to the pigment opsin to form rhodopsin in the plasma membrane of rod cells of the retina. These cells are responsible for vision in low light. Similar complexes occur in the cone cells of the retina, which are responsible for color vision and vision in bright light. When light of the correct wavelength hits a rod or cone cell, rhodopsin undergoes a change in conformation, which results in an action potential (change in voltage across the cell membrane). This action potential is then transmitted via neurons to the brain, giving rise to the perception of sight. Vitamin A has a number of other roles in human metabolism, but clinical symptoms of VADD are usually related to vision; this is not surprising, considering the central importance of vitamin A to vision.

Plant Biotechnology

117

Beta-carotene

Retinal

H C O

FIGURE 4.13 Structure of β-carotene and retinal.

The continuing and widespread nature of VADD has led people to investigate a number of potential solutions. One option is to develop staple foods that have increased levels of β-carotene. The close link between rice dependency and VADD led researchers in Switzerland and Germany to develop golden rice. The research group, led by Ingo Potrykus and Peter Beyer, with funding from the Swiss government and the Rockefeller Foundation, developed a transgenic line of rice with increased levels of β-carotene. Because of the resulting yellow color in the endosperm, these seeds were dubbed “golden rice.” It was necessary to use a transgenic approach rather than classical plant breeding techniques because rice does not normally have β-carotene in the inner endosperm (the area of the cereal seed that stores starches and other nutrients). It is essential that the provitamin be stored in the endosperm because this constitutes the bulk of the rice grain after milling. Rice actually contains substantial amounts of β-carotene in the aleurone layer of the endosperm, but this oil-rich layer is removed by milling. If it is not removed, the rice has a greatly increased risk of developing rancidity, particularly in tropical regions. The development of golden rice posed numerous technical difficulties. β-Carotene in rice is produced from a pool of geranyl geranyl diphosphate (GGPP), an isoprenoid that is used to form a variety of compounds. Three enzymes are necessary to convert GGPP to β-carotene; none of these enzymes are active in the rice endosperm. Introducing three genes simultaneously is much more difficult than introducing a single gene into a plant. Potrykus et al. used several agrobacteriumbased approaches; we will focus on one of these. They inserted genes for two of the enzymes (psy, coding for phytoene synthase, and crt, coding for phytoene desaturase) into one plasmid vector (pZPsC), and the other gene (lcy, coding for lycopene β-cyclase) into a separate plasmid (pZLcyH) that also had a selectable marker gene (aph IV, coding for hygromycin resistance). In each of these plasmids (Figure 4.14), left and right borders were present, allowing transfer of the genes by vir genes of A. tumefaciens. Two of the genes (psy and lcy) were preceded by endosperm-specific promoters (i.e., the genes would be expressed only in the

118

Introduction to Food Biotechnology

p2

p2

crt aph IV

psy rb

p1

lb

p1

lb

rb

ori A. tumefaciens pZPsC

lcy

ori A. tumefaciens pZLcyH

FIGURE 4.14 Structure of pZPsC and pZLcyH, the vectors used in the development of β-carotene-enriched rice (golden rice). Promoters are indicated by p1 (endosperm-specific promoter) and p2 (CaMV 35S promoter). Arrows indicate the direction of transcription from the promoters. (Modified from Ye X., et al., Science, 287, 303, 2000.)

endosperm); crt and aph IV were preceded by the CaMV 35S promoter, which induces constitutive (all tissues, all times) gene expression. The researchers introduced the two vectors separately into A. tumefaciens by electroporation. The two resulting transformed bacterial strains were then co-incubated with immature rice embryos (these are produced through plant tissue culture methods). Transformed plant cells were then selected using hygromycin, which killed nontransformed plant cells. Plants were then regenerated from callus cultures of cells that survived the hygromycin treatment. This co-transformation was a clever approach, because it turned out that lycopene cyclase is not required for the production of β-carotene in rice endosperm (it is probably constitutively produced). Because the vector that contained the lcy gene also contained the selectable marker gene for hygromycin resistance, it should be possible to remove the marker gene through traditional breeding methods, without losing β-carotene production. Genes from the two plasmids were probably inserted at widely different locations in the plant cell’s chromosomes. Transformed plants were grown to maturity and seeds were collected. Not all of the seeds were β-carotene enriched, but in a typical line, the average β-carotene content was 1.6 µg/g of endosperm tissue. Considering that this β-carotene measure was based on a mixture of β-carotene and non-β-carotene-enriched seeds, Potrykus et al. postulated that pure breeding lines (i.e., lines that produced 100% β-caroteneenriched seeds) would contain at least 2 µg/g β-carotene. This could correspond to 100 µg of retinol equivalents in a daily serving of 300 g of golden rice, possibly enough provitamin A to alleviate vitamin A deficiency. The publication of the golden rice manuscript provoked a flurry of commentary from both pro- and antibiotechnology forces. The beleaguered corporate food biotechnology community, beset by widespread criticisms of arrogance (forcing transgenic

Plant Biotechnology

119

crops onto the world), soon began to use golden rice as a poster child to promote transgenic technology. The antibiotechnology forces, led by Greenpeace, denounced golden rice, maintaining that more than 7 lb of dry golden rice would have to be consumed daily for an individual to gain enough provitamin A. At this time, it is difficult to assess the utility of golden rice in decreasing the incidence of VADD, primarily because of uncertainties about the bioavailability of β-carotene in golden rice. It is well known that simply measuring β-carotene content of a fruit or vegetable does not give an accurate picture of the ability to deliver β-carotene after ingestion. Many factors affect this; the matrix surrounding β-carotene is particularly important. For example, many leafy green vegetables have high levels of β-carotene, but it is associated with protein complexes in chloroplasts. These complexes are poorly digested in the human gut, resulting in poor intake of β-carotene. Several other factors add to the uncertainty. Required levels of provitamin A vary widely, depending on age and numerous other factors. Thus, the optimal level of intake is difficult to assess. Also, human absorption of β-carotene is linked to fat consumption; people with extremely low levels of fat intake (95% of the tocopherol being in the form of α-tocopherol. It is expected that this approach would also be successful in oilseed crops such as soybean and canola, leading to vitamin E–enhanced transgenic oilseeds. Oilseeds are also promising candidates for development of transgenics with increased levels of health-promoting fatty acids. Vegetable oils are the main source of PUFAs in most populations. Given the strong interest in the health benefits of certain PUFA, and health risks associated with other PUFA, it is not surprising that several research groups are interested in using transgenic technology to modify the fatty acid profile of oilseeds. For example, it is becoming clear that it would be desirable to have vegetable oils containing large amounts of linolenic acid and smaller amounts of linoleic acid. This could conceivably be done by changing the degree of expression of the enzymes responsible for making these fatty acids. Such an approach has been used to create high-laurate canola. However, this was done not to increase the healthfulness of the oil, but because of the many industrial uses of laurate (e.g., in detergent manufacture). It is expected, though, that in the near future vegetable oils with increased health functionality will be available.

H. REDUCING

OR

ELIMINATING ALLERGENS

For reasons that are unclear, the incidence of pollen-associated allergies in many countries appears to be increasing. Although less well documented, it is also likely that food allergies are becoming more frequent. However, there are many misconceptions among the media and the general population about the nature and extent of food allergies. The formal definition of a food allergy is an adverse reaction to food that is caused by an immune reaction. Many reactions commonly assumed to be allergic reactions are actually food intolerance, which is often caused by specific enzyme (e.g., lactase) deficiency. True food allergy appears to be relatively uncommon. A recent study of food allergy perception and incidence in the Netherlands found that 12.4% of 1483 questionnaire respondents indicated that they suffered from food allergies. These people were then tested through double-blind, placebocontrolled food challenges. The incidence of true food allergy in this study was only 0.8%. The reasons for the discrepancy between perceived and clinical food allergy are probably related to the popular media and the prevalence of alternative medical practitioners who blame food allergies for a host of medical problems (e.g., complaints of poor general health). For some people, however, food allergy is a matter of life and death. Peanut allergies, for example, sometimes lead to life-threatening anaphylactic reactions. For

130

Introduction to Food Biotechnology

these people, as well as for those with milder food allergies, it is desirable to reduce the levels of allergens (compounds that provoke an allergic reaction) in foods. It is difficult to assess the allergenicity of different foods because susceptibility among humans varies with several factors, with age probably the most important. However, common culprits are eggs, fish, celery, carrots, kiwi, cow’s milk, legumes, and sunflower seeds. Because it is much easier to transform plants than animals, most discussions of allergen reduction have focused on vegetables. The best example to date of allergen reduction is in transgenic rice lines that have reduced levels of a 16-kDa (kilodalton, a measure of molecular mass) allergenic protein. This protein was assumed to have some allergenicity because it reacted with antibodies in serum collected from people with rice allergy. The gene for this protein was cloned by the following procedure: • mRNA was collected from developing rice seeds, and a cDNA library was established using a phage vector (γgt11). • This library was screened using a nucleic acid probe that was based on the amino acid sequence of one end of the 16-kDa protein. • A DNA library was then probed to isolate the promoter region of the 16-kDa gene. • A plasmid vector (similar to pUC) was constructed that contained the 16-kDa promoter in front of the antisense sequence of the 16-kDa gene. • This plasmid was introduced into rice protoplasts via electroporation; transformed cells were selected through use of an antibiotic marker gene. • Rice plants were then regenerated from the protoplasts. Transformed rice plants produced seeds that had about one fifth the amount of allergenic protein as untransformed plants. It is uncertain whether this would be enough of a reduction to prevent reactions among allergenic individuals. It is also possible that other proteins act as allergens, in addition to the 16-kDa protein. The function of the 16-kDa protein is also uncertain; it has sequence similarity with barley trypsin inhibitor and wheat α-amylase inhibitor, and therefore may function as an antiherbivore compound. This would imply that changing the sequence of the 16-kDa protein might improve rice digestibility, in addition to reducing allergenicity. Nonetheless, it is essential to learn more about the normal function of this protein, to assess the ramifications of using antisense technology to modify its expression. Other aspects of transgenic technology may be relevant to food allergies. For example, it has been demonstrated that milk proteins can be expressed in plant tissue; perhaps we could engineer a plant that produces proteins with the nutritional quality of human milk but without the allergenicity associated with cow’s milk. Of course, food allergy is important to transgenic technology in another context — regulatory bodies usually carefully consider the potential allergenicity of transgenic proteins before allowing their release and consumption. This important point will be further discussed in Chapter 9.

Plant Biotechnology

131

VI. ALTERNATIVE AGRICULTURAL TECHNOLOGIES Increasing numbers of growers and corporations are growing and marketing foods that appeal to consumers who are in favor of alternative farming methods, especially those that are less reliant on the use of agrochemicals. Organic agriculture is a striking example of how shifts in consumer attitudes can change the food industry. Twenty years ago organic farming operated at a loosely organized small scale, catering to a small but devoted following. Now, the organic food industry contributes billions of dollars to economies throughout the world. It is no longer solely the domain of small market gardeners, and it is becoming increasingly influenced by large corporations. Although this process may be contrary to the beliefs of many proponents of organic agriculture, it has allowed prices to decrease somewhat, which in turn has expanded the organic food consumer base. However, there still exists a pronounced price difference between most types of organic food and conventional food. Less extreme modifications of conventional agriculture (use of biofertilizers or integrated pest management [IPM]) may appeal to consumers who are not willing to pay the premium price of organic food but perceive conventional agriculture as overdependent on toxic chemicals. This could be a marketing opportunity for both producers and food companies. Consequently, it is useful to be aware of some of the technologies that benefit both organic farmers and farmers interested in reducing their use of agrochemicals. Organic farmers do not use synthetic pesticides to control weeds and pests. They also foreswear chemical fertilizers. Instead, weeds are controlled physically through tilling and other methods of disrupting weed growth. Insect pests are usually controlled through a combination of “natural” insecticides (e.g., rotenone), detergents (insecticidal soap), or microbial insecticides (usually spores of B. thuringiensis). It is ironic that consumers of organic food have such intense antipathy to food biotechnology, because many organic growers are highly dependent on the ability of microbial biotechnologists to cheaply grow large amounts of B. thuringiensis spores. Transgenic crops with Bt genes would also be useful tools for organic growers to control insect pests, but proponents of organic agriculture generally reject any use of recombinant DNA technology. Chemical fertilizers are replaced by either animal or green manures. Green manures are crops grown on soil to accumulate nutrient-rich organic matter. Legumes are a popular green manure because of their ability to form symbioses with soil bacteria that allow the plants to obtain nitrogen from the atmosphere. Consequently, leguminous green manures can add substantial amounts of nitrogen to a soil. Because nitrogen is the element (after carbon, hydrogen, and oxygen) most required by plants, this is an important consideration. Biotechnology has much to offer organic farmers — to improve mineral supply to plants and to help control pests. It is beyond the scope of this book to address all aspects of biological pest control (the use of microbes or other biological agents to control pests), but we will briefly examine biofertilizers. Biofertilizers improve the ability of plants to obtain mineral nutrients from the soil. This technology exploits the natural abilities of microbes such as Rhizobium,

132

Introduction to Food Biotechnology

a bacterium that can form symbioses with the roots of legumes. Rhizobium is able to fix nitrogen from the atmosphere into ammonium, which can then be used by the plant as a source of nitrogen. Most soils naturally contain Rhizobium, but naturally occurring strains are often inefficient fixers of nitrogen. Manipulation of Rhizobium populations can correct this problem. The use of efficient Rhizobium strains in the cultivation of peas, beans, and other legumes allows decreased reliance on chemical fertilizers. Application of nitrogenous fertilizers is often a large economic burden for farmers, and it can lead to problems such as increased rates of erosion and contamination of underground reservoirs with nitrate. This is a problem for users of underground reservoirs because of hazards associated with the consumption of nitrate in drinking water. Biofertilizers can also increase the supply of phosphorus to plants. Mycorrhizal fungi are extremely common symbionts of plant roots, and they improve plant growth and health through a number of mechanisms. Increased rate of phosphorus supply is usually the most clear-cut benefit to plants from mycorrhizas. Specific groups of soil bacteria can also improve the rate of phosphorus intake by plants. These are the phosphate-solubilizing bacteria; their growth and activity lead to increased levels of phosphate in the soil solution. Both of these groups of microorganisms are normally present in agricultural soils, but, as with the case of Rhizobium, the most efficient species are not always present, and some soils have low numbers of mycorrhizal fungi and phosphate-solubilizing bacteria. IPM is similar in some ways to organic agriculture because of its decreased reliance on chemical pesticides and increased reliance on biological control of pests (e.g., using a specific fungal pathogen to control a specific weed). IPM also involves the use of assessment methods that keep track of insects and other pests throughout the growing season. This allows the judicious use of chemical pesticides — they are applied only when a particular pest has reached levels that justify the expense. Consumers throughout the world rate pesticide residues as a major concern, despite reassurances from government regulators and the chemical industry that current pesticide use is safe. Consequently, it represents another potential marketing niche that may be exploited more in the future, if consumer concerns over food safety continue to increase. However, this might pose legal and regulatory problems, because most countries have regulations governing the use of “organic” labels but have no mechanism for labeling food that is produced through reduced use of agrochemicals (chemical fertilizers and pesticides). This is important because many farmers find the requirements for organic certification daunting (e.g., a number of agrochemical-free years of production are typically required before a farmer can claim organic status). Alternatives to organic labels could be useful to such farmers. Unfortunately, IPM and biofertilizers have received little media attention, and the public remains largely unaware of biotechnologies that have the potential to decrease agriculture’s reliance on agrochemicals. Similarly, despite many attempts by food biotechnologists to justify the use of transgenic technology based on its potential to decrease pesticide and fertilizer use, most consumers appear unconvinced. The next few years are crucial to plant biotechnologists; they need to demonstrate that plant biotechnology is safe to consumers and can benefit consumers

Plant Biotechnology

133

directly, via improved food functionality, for example, and indirectly, through decreased use of agrochemicals that are detrimental to the environment.

RECOMMENDED READING 1. Wilkinson, J., Biotech plants: from lab bench to supermarket shelf, Food Technol., 51, 37, 1997. 2. Dunwell, J. M., Transgenic crops: the next generation, or an example of 2020 vision, Ann. Bot., 84, 269, 1999. 3. Raven, P. H., Evert, R. F., and Eichhorn, S. E., Biology of Plants, 6th ed., Freeman, New York, 1999. 4. Wenzel, G., Application of unconventional techniques in classical plant production, in Plant Biotechnology, Fowler, M. W. and Warren, G. S., Eds., Pergamon Press, Oxford, 1992, chap. 13. 5. Knorr, D., Caster, C., Dörneburg, H., Dorn, R., Gräf, S., Havkin-Frenkel, D., Podstolski, A., and Werrman, U., Biosynthesis and yield improvement of food ingredients from plant cell and tissue cultures, Food Technol., 77, 57, 1993. 6. Hall, R. D., An introduction to plant cell culture: pointers to success, in Plant Cell Culture Protocols, Hall, R. D., Ed., Agritech, Shrub Oak, NY, 1999, chap. 1. 7. Zupan, J., Muth, T. R., Draper, O., and Zambryski, P., The transfer of DNA from Agrobacterium tumefaciens into plants: a feast of fundamental insights, Plant J., 23, 11, 2000. 8. McKinnon, G. E. and Henry, R. J., Control of gene expression for the genetic engineering of cereal quality, J. Cereal Sci., 22, 203, 1995. 9. Koziel, M. G., Beland, G. L., Bowman, C., Carozzi, N. B., Crenshaw, R., Crossland, L., Dawson, J., Desai, N., Hill, M., Kadwell, S., Launis, K., Lewis, K., Maddox, D., McPherson, K., Meghji, M. R., Merlin, E., Rhodes, R., Warren, G. W., Wright, M., and Evola, S. V., Field performance of elite transgenic maize plants expressing an insecticidal protein derived from Bacillus thuringiensis, Bio/Technology, 11, 194, 1993. 10. Armstron, C. L., Parker, G. B., Pershing, J. C., Brown, S. M., Sanders, P. R., Duncan, D. R., Stone, T., Dean, D. A., DeBoer, D. L., Hart, J., Howe, A. R., Morrish, F. M., Pajeau, M. E., Petersen, W. L., Reich, B. J., Rodriguez, R., Santino, C. G., Sato, S. J., Schuler, W., Sims, S. R., Stehling, S., Tarochione, L. J., and Fromm, M. E., Field evaluation of European corn borer control in progeny of 173 transgenic corn events expressing an insecticidal protein from Bacillus thuringiensis, Crop Sci., 35, 550, 1995. 11. Hansen Jesse, L. C. and Obrycki, J. J., Field deposition of Bt transgenic corn pollen: lethal effects on the monarch butterfly, Oecologia, 125, 241, 2000. 12. Losey, J. E., Rayor, L. S., and Carter, M. E., Transgenic pollen harms monarch larvae, Nature, 399, 214, 1999. 13. Wraight, C. L., Zangeri, A. R., Carroll, M. J., and Berenbaum, M. R., Absence of toxicity of Bacillus thuringiensis pollen to black swallowtails under field conditions, Proc. Natl. Acad. Sci. U.S.A., 94, 770, 2000. 14. Gould, F., Sustainability of transgenic insecticidal cultivars: integrating pest genetics and ecology, Annu. Rev. Entomol., 43, 701, 1998. 15. Hilder, V. A. and Boulter, D., Genetic engineering of crop plants for insect resistance — a critical review, Crop Prot., 18, 177, 1999.

134

Introduction to Food Biotechnology

16. Munkvold, G. P., Hellmich, R. L., and Rice, L. G., Comparison of fumonisin concentrations in kernels of transgenic Bt maize hybrids and nontransgenic hybrids, Plant Dis., 83, 130, 1999. 17. Gressel, J., Molecular biology of weed control, Transgenic Res., 9, 355, 2000. 18. Schuch, W., Improving tomato quality through biotechnology, Food Technol., November, 78–83, 1994. 19. Turner, M. and Schuch, W., Post-transcriptional gene-silencing and RNA interference: genetic immunity, mechanisms and applications, J. Chem. Technol. Biotechnol., 75, 869, 2000. 20. Madhavi, D. L. and Salunkhe, D. K., Tomato, in Handbook of Vegetable Science and Technology, Salunkhe, D. K. and Kadam, S. S., Eds., Marcel Dekker, New York, 1998, chap. 7. 21. Redenbaugh, K., Hiatt, W., Martineau, B., and Emlay, E., Regulatory assessment of the FLAVR SAVR tomato, Trans. Food Sci. Technol., 5, 105, 1994. 22. McLaren, D. and Frigg, M., Sight and Life Manual on Vitamin A Deficiency Disorders (VADD), 2nd ed., Task Force Sight and Life, Basel, 2001. 23. Guerinot, M. L., The green revolution strikes gold, Science, 287, 241, 2000. 24. Ye, X., Al-Babili, S., Klöti, A., Zhang, J., Lucca, P., Beyer, P., and Potrykus, I., Engineering the provitamin A (β-carotene) biosynthetic pathway into (carotenoidfree) rice endosperm, Science, 287, 303, 2000. 25. Waterhouse, P. M., Wang, M.-G., and Lough, T., Gene silencing as an adaptive defence against viruses, Nature, 411, 834, 2001. 26. Stuiver, M. H. and Custers, J. H. H. V., Engineering disease resistance in plants, Nature, 411, 86, 2001. 27. Lorito, M. and Scala, F., Microbial genes expressed in transgenic plants to improve disease resistance, J. Plant Pathol., 81, 73, 1999. 28. Vandemark, G. J., Transgenic plants for the improvement of field characteristics limiting crop production, in Molecular Biotechnology for Plant Food Production, Paredes-López, O., Ed., Technomic, Lancaster, PA, 1999, Chap. 6. 29. Guzmán-Maldonado, S. H. and Paredes-López, O., Biotechnology for the improvement of nutritional quality of food crop plants, in Molecular Biotechnology for Plant Food Production, Paredes-López, O., Ed., Technomic, Lancaster, PA, 1999, chap. 14. 30. Duranti, M. and Gius, C., Legume seeds: protein content and nutritional value, Field Crops Res., 53, 31, 1997. 31. Pen, J., Verwoer, T. C., van Paridon, P. A., Beudeker, R. F., van den Elzen, P. J. M., Geerse, K., van der Klis, J. D., Versteegh, H. A. J., van Ooyen, A. J. J., and Hoekema, A., Phytase-containing transgenic seeds as a novel feed additive for improved phosphorus utilization, Bio/Technology, 11, 811, 1993. 32. Raboy, V., Low-phytic-acid grains, Food Nutr. Bull., 21, 423, 2000. 33. Davies, C. S., Nielsen, S. S., and Nielsen, N. C., Flavor improvement of soybean preparations by genetic removal of lipoxygenase-2, J. Am. Oil Chem. Soc., 64, 1428, 1987. 34. Utsumi, S., Plant food protein engineering, Adv. Food Nutr. Res., 36, 89, 1992. 35. Schulman, A., Chemistry, biosynthesis, and engineering of starches and other carbohydrates, in Molecular Biotechnology for Plant Food Production, Paredes-López, O., Ed., Technomic, Lancaster, PA, 1999, chap. 12. 36. Schuurink, R. C. and Louwerse, J. D., The genetic transformation of wheat and barley, in Cereal Biotechnology, Morris, P. C. and Bryce, J. H., Eds., Woodhead Publishing, Cambridge, 2000, chap. 2.

Plant Biotechnology

135

37. Del Vecchio, A. J., High laurate canola: how Calgene’s program began, where it’s headed, Inform, 7, 230, 1996. 38. Henry, R. J., Using biotechnology to add value to cereals, in Cereal Biotechnology, Morris, P. C. and Bryce, J. H., Eds., Woodhead Publishing, Cambridge, 2000, chap. 5. 39. Faus, I., Recent developments in the characterization and biotechnological production of sweet-tasting proteins, Appl. Microbiol. Biotechnol., 53, 145, 2000. 40. Shintani, D. and DellaPenna, D., Elevating the vitamin E content of plants through metabolic engineering, Science, 282, 2098, 1998. 41. Davey, M. W., van Montagu, M., Inzé, D., Sanmarin, M., Kanellis, A., Smirnoff, N., Benzie, I. J. J., Strain, J. J., Favell, D., and Fletcher, J., Plant L-ascorbic acid: chemistry, function, metabolism, bioavailability and effects of processing, J. Sci. Food Agric., 80, 825, 2000. 42. Matsuda, T., Nakase, M., Adachi, T., Nakamura, R., Tada, Y., Shimada, H., Takahashi, M., and Fujimara, T., Allergenic proteins in rice: strategies for reduction and evaluation, in Food Allergies and Intolerances, Eisenbrand, G., Aulepp, H., Dayan, D. D., Elias, P. S., Grunow, W., Ring, J. and Schlatter, J., Eds., VCH Publishers, Weinheim, Germany, 1996, chap. 12. 43. Wüthrich, B., Epidemiology of allergies and intolerances caused by foods and food additives: the problem of data validity, in Food Allergies and Intolerances, Eisenbrand, G., Aulepp, H., Dayan, D. D., Elias, P. S., Grunow, W., Ring, J., and Schlatter, J., Eds., VCH Publishers, Weinheim, Germany, 1996, chap. 1. 44. Coleman, E., The New Organic Grower, Nimbus, Halifax, Nova Scotia, Canada, 1995.

5

Animal Biotechnology I. OVERVIEW

On the whole, humans are quite fond of eating meat, fish, and dairy products. In the industrialized world, consumption of some animal products (e.g., red meat) has declined, but this has frequently been balanced by increased consumption of others (e.g., poultry). In the developing world, meat consumption is often viewed as one of the prime benefits of increased personal wealth. Most animal products are excellent, balanced sources of protein and supply vitamins and minerals such as iron. Thus they are considered to be an essential part of the diet by most nutritionists. Animal products can be made into a diverse array of foods, encompassing fresh, cured, and precooked products; this process is a potent and vibrant part of the global economy. Why does meat eating persist, despite the well-known health risks associated with diets with high levels of animal products (especially those that have a high fat content) and the environmental problems associated with animal production (e.g., contamination of water sources by manure from feedlot operations)? The answer is probably partly hard-wired into our genes — humans like to eat meat. The common salivary response to the odor of barbecued meat is perhaps an illustration of this phenomenom. A diverse range of animals are used as food. Mammalian food animals are divided into two categories: (1) ruminants (cattle, goats, sheep) — these animals have a rumen, a specialized stomach compartment that incubates ingested plant material for a protracted period of time before allowing it to pass to the intestine; and (2) monogastrics (pigs), animals that have only a single stomach compartment. Ruminants rely on bacteria and other microorganisms in their rumen for digestion of cellulose-rich plant biomass. The rumen is composed of a complex, interdependent community of microbes that converts indigestible (to the animal) plant feed into volatile fatty acids (e.g., acetate), which are then absorbed by the ruminant and used as a source of energy and carbon. In addition to these mammalian food animals, several species of birds (chickens, turkeys, ducks, and geese) and their eggs are important animal products. Fish are unique among food animals in that wild fish are more important than domestic fish (aquaculture); however, aquaculture is rapidly increasing in popularity, partly because of depletion of global fish stocks. Overfishing, environmental degradation (e.g., increased flow of industrial and domestic sewage to the ocean), and climatic shifts are responsible for the precipitous decrease in certain fish stocks. Biotechnology plays a role in the production of meat, eggs, dairy products, and fish. Microbial biotechnology is very useful, particularly in the production of fermented dairy products; this technology will be discussed in Chapter 7. Diagnostic

137

138

Introduction to Food Biotechnology

biotechnology that detects and helps diagnose pathogenic bacteria is increasingly important in the slaughterhouse environment because of the increasing incidence of cattle carrying dangerous bacteria such as Escherichia coli O157:H7. Diagnostic applications will be discussed in Chapter 6. Here we will concentrate on the potential of transgenic technology to change food animals. Transgenic feed crops are also relevant to animal production (e.g., transgenic soybeans with increased phytase; see Chapter 4, Section V.C). Transgenic animals are not currently used by the food industry, but this will likely change soon. The first commercially grown transgenic animal will probably be salmon with boosted levels of growth hormone. These fish will be restricted to the aquaculture industry, and their success hinges partly on regulatory approval and partly on consumer acceptance. Given the outcry against transgenic plants, transgenic fish will likely be met with substantial public hostility, especially from antibiotechnology activist groups. The lack of transgenic animals in the food industry is at first glance surprising, given the enormous success of transgenic animals in elucidating many aspects of the molecular biology of animals. Transgenic mice in particular have been useful research tools because they have allowed scientists to increase, decrease, or eliminate (“knockout”) specific genes and then examine the effects of the gene modification on the animal’s physiology or development. However, when a biotechnologist begins research leading to the production of a transgenic animal destined for agricultural use, the research program becomes very different. The end product of the research must be an animal that is improved in some way, in comparison to conventional animals. This improvement must not be accompanied by adverse effects on the health of the animal or on the safety of foods derived from the animal. Another daunting factor is cost; generally it is more expensive to produce transgenic animals than transgenic plants, and this cost rapidly increases with the size of the animal. Indeed, biotechnology companies have shied away from funding research into transgenic farm animals; most researchers in this field are funded by government organizations. Finally, the biotechnologist is virtually guaranteed to face considerable public opposition and regulatory hurdles after spending years developing the transgenic animal. So, perhaps it is not surprising that few laboratories are currently involved in research and development of transgenic animals for the food industry. Despite this pessimistic outlook, though, transgenic biotechnology has promising applications (Table 5.1). The first three traits in Table 5.1 will be discussed later in this chapter. Control of sex ratios in poultry is desirable because producers of broiler chickens use only male chicks and discard female chicks. In contrast, egg producers use only female chicks. This large-scale wastage could be avoided if producers could control the sex ratio. Unfortunately, poultry scientists have not yet deciphered the mechanisms involved in control of sex ratios, making it impossible to develop transgenic chickens with useful modifications of this phenotype. Many animal scientists hope that transgenic technology will be useful in introducing disease resistance to ruminants, monogastrics, and poultry. This may be an effective approach to control viruses (e.g., the virus that causes foot-and-mouth disease). Many animal viruses are difficult to control using conventional methods (e.g., vaccination), but could be controlled by changing the animal’s genome so that

Animal Biotechnology

139

TABLE 5.1 Potential Uses (and Problems) of Transgenic Biotechnology in the Production of Animal-Based Foods Animal

Transgene

Objective

Problems

Ruminants

Growth hormone

Many side effects

Ruminants

Modified milk proteins

Fish Poultry Various animals

Growth hormone ? Pathogen receptors

Leaner meat, increased feed efficiency Improved processing, improved nutrition Faster growth Control over sex-ratios Disease resistance

Technically difficult Environmental issues Unknown genes are involved Technically difficult; may carry numerous side effects

it can no longer be infected. For example, viruses get into host cells through binding to specific receptors. If these receptors (usually proteins embedded in the host cell’s membrane) could be changed or removed, the virus would be unable to infect the cell. Unfortunately, it is unlikely that this could be done without adversely affecting the animal’s metabolism or development — cell surface proteins always have a function of some sort. This chapter concentrates on the two applications of transgenic technology that are most likely to have an impact on food production in the foreseeable future: fish with enhanced production of growth hormones and ruminants with modified milk proteins. The last two sections address alternative approaches to the development of transgenic animals: the use of embryonic stem cells and the cloning of adult cells.

II. TRANSGENIC FISH The most famous transgenic animal is the “supermouse,” a mouse that had increased expression of growth hormone; these mice, developed in the early 1980s, were twice as large as their nontransgenic littermates. The success of this research likely fostered expectations that farm animals would soon be developed that had boosted levels of growth hormone. Such animals might have leaner meat, better feed efficiency (ability to convert feed biomass into animal biomass), and faster growth. Indeed, transgenic pigs have these improvements, but they are also susceptible to a number of abnormalities, particularly in skeletal development. These problems are likely related to the increased expression of growth hormone, and not to other factors (e.g., the process used to create the transgenic animals), because similar abnormalities can be induced through direct application of growth hormone to animals such as pigs. Because of these serious health problems, transgenic farm animals with increased levels of growth hormone have not yet been released. Interestingly, this has not stopped some meat processors in North America from attempting to label their food as “not genetically modified,” a non sequitur because currently there are no commercially available transgenic food animals.

140

Introduction to Food Biotechnology

Biotechnologists have had much greater success in the development of transgenic fish with boosted levels of growth hormone, although success has been largely limited to salmonid fish (salmon and trout are salmonids important to the food industry). The rationale for these transgenics is that production efficiency is the major variable affecting the salmon aquaculture industry. Efficiency is determined largely by feed efficiency but also by growth rates. It has long been known that applying fish growth hormone to salmon results in faster growth and increased feed efficiency, so producing transgenic fish with higher levels of endogenous (from within) growth hormone was a logical step. Salmon are also difficult to rear in aquaculture because of the two distinct phases of their life cycle — initially, they require cold freshwater (which leads to slow growth rates) or heated freshwater (which is expensive), and after several years of this juvenile rearing, they need to be transferred to saltwater. Increased levels of growth hormone shorten the juvenile rearing period, thus cutting costs of production. The overall strategy of creating a transgenic animal is similar to that used to develop a transgenic plant (Chapter 4, Section III.B). The researcher clones the gene of interest and finds out as much as is logistically possible about the gene. For example, knowledge of the DNA sequence that codes for the desired protein and an understanding of promoters, terminators, and regulatory regions of the gene are useful. Promoters and other genetic elements are then chosen and linked to the desired gene. This DNA is then linearized and any other genes (e.g., antibiotic selectable marker genes) are removed. With fish, the linear DNA is then injected into the nuclei of cells in just-fertilized egg cells (with fish, injection into the cytoplasm of the fertilized egg cell is also likely to be successful). The injected DNA will occasionally become permanently integrated into one of the cell’s chromosomes. With most animals, this is a rare event and is difficult to control, because the integration mechanism is unknown. Consequently, many attempts at micro-injection of egg cells end in failure. Micro-injection is relatively straightforward with salmonid eggs, partly because large numbers of eggs can easily be obtained, either manually or with the mechanical assistance of a micromanipulator — a device that gives the operator fine-tuned control over the injection process. Note that it is not necessary to integrate desired genes into vectors (e.g., plasmids) before micro-injection into nuclei. For this reason, the injected DNA, along with its promoter, is called a construct. Once the embryo develops into an adult, the inserted gene is likely to be active only in certain cells, dictated by the chromosomal location of insertion (different regions of the genome are actively expressed in different cell types). However, the trait should be inheritable, and pure-breeding transgenic lines can usually be obtained after several generations. A number of lessons were learned during early attempts to generate transgenic salmonids with increased levels of growth hormone. First, biotechnologists used mammalian growth hormone genes and mammalian promoters. In most cases, this did not lead to increased production of growth hormone. The reason for this is unclear, but it may be related to incorrect mRNA processing due to the lack of introns. Molecular biologists have observed that transgenic plants and animals sometimes produce larger amounts of heterologous protein when introns are present in the transgene.

Animal Biotechnology

141

When growth hormone genes from fish are inserted into fish with strong promoters, efficient expression (and, presumably, correct mRNA processing) results. It has also become clear that it is preferable to use promoters derived from fish rather than mammalian promoters. Promoters that demonstrably drive growth hormone expression in transgenic salmon include the ocean pout antifreeze protein promoter and the sockeye salmon metallothionein-B promoter. These promoters are active in the liver; this constitutes a major change from native growth hormone production, which occurs only in the pituitary gland. Transgenic fish with increased growth hormone grow faster and accumulate biomass more efficiently than conventional fish. However, the technology is not perfect. Depending on the level of transgene expression, some fish may develop acromegaly, a condition involving excessive growth of the cartilage in the head region. This can lead to decreased fish viability in extreme cases. When growth hormone is expressed at moderate levels, acromegaly is not seen, but this is accompanied by a more restrained increase in growth rates. Before being released for commercial production, this problem must be solved. Regulators must also consider environmental risks before allowing release of transgenic salmon. Many environmentalists and ecologists fear that transgenic fish will escape from their aquaculture cages, with unpredictable consequences. A common fear expressed by antibiotechnology activists is that transgenic fish will outcompete their wild counterparts, thus reducing fish diversity. However, most fish ecologists have the opposite concern, predicting that transgenic fish will be relatively unfit compared to their wild counterparts. In this context, “fitness” refers to the overall ability of an individual to reproduce successfully. Even if transgenic fish have reduced fitness, they could still exert harmful effects. If they interbreed with wild salmon, they might introduce deleterious genes into wild salmon populations. This “trojan gene” effect could occur if transgenic fish are more successful at mating than wild fish but have decreased viability. Computer simulations have indicated that this could lead to a decline in wild populations. Transgenic fish with boosted levels of growth hormone may have better fitness than wild salmon, leading to displacement of populations of wild salmon. This is a sensitive issue, because many salmon populations, particularly those of Atlantic salmon, are endangered; any extra stress imposed by escaped transgenic salmon would be undesirable. The major problem with risk assessment of transgenic fish is that the interactions among specific genotypes, fitness, and the environment are poorly understood. Risk assessment is also complicated by the large number of phenotypes that change (Table 5.2) upon the addition of a single gene construct (gene + promoter). Some of these phenotypes would seem to give transgenic fish a competitive advantage over normal fish; for example, increased feeding motivation seems to be a positive trait. However, increased feeding motivation, combined with decreased swimming ability, might make transgenic fish more susceptible to predation. Those for and against fish biotechnology agree on one point: transgenic fish must be physically and biologically contained, to minimize escapes and the consequences of escapes. Physical containment is achieved through the use of cages and nets, and biological containment can be done either by sterilizing fish chemically

142

Introduction to Food Biotechnology

TABLE 5.2 Phenotypes of Salmonids that Change upon Introduction of a Growth Hormone Gene Phenotype

Nature of Change

Growth rate Smoltification Appetite and feeding motivation Metabolism Swimming ability Cranial morphology Muscle structure Life cycle

Increased Earlier Increased Increased rate Decreased Abnormal Increased hyperplasia Shortened

or by clever breeding manipulation that produces sterile triploid fish (see item 6 in R. O’Flynn et al., 1997, for an explanation of this procedure). It is important to note that salmon or other fish that have been selectively bred for rapid growth rates using traditional breeding techniques carry similar risks upon escape into the environment. But it is virtually impossible to “follow” them after such an escape; in contrast, transgenic fish carry unique genetic sequences that can be identified through PCR (polymerase chain reaction) and other techniques. The final point that transgenic fish developers must address is consumer acceptance. The only potential benefit to the consumer of transgenic fish with enhanced growth rates is price. This may not be enough to counter consumer resistance, particularly in the face of vigorous action by antibiotechnology forces, a virtual certainty in the case of transgenic fish.

III. MODIFIED MILK PROTEINS In the early 1980s, it became clear that it was technically possible to transfer specific genes to mammals using recombinant DNA technology. This led to numerous attempts to introduce new or modified genes into farm animals. With some species (e.g., poultry), transgenic technology is now well established in the laboratory. However, it has proven difficult to produce transgenic animals in cattle, the most important food animal. There are several reasons for this: (1) cattle have long gestation periods, (2) they usually produce only one calf per gestation, and (3) zygotes and embryos of cattle are more difficult to manipulate than those of other farm animals. These factors make it expensive to produce transgenic cattle — large herd sizes are necessary in order to produce enough viable transgenic individuals. Because of these technical problems, most research into transgenic cattle has focused on the production of human proteins that can be used to treat human disease. This molecular farming (“pharming”) has the potential to be lucrative for pharmaceutical companies. In some cases, one or two transgenic cows could conceivably produce enough therapeutic protein to serve a $100 million market.

Animal Biotechnology

143

TABLE 5.3 Milk Traits that Could Be Modified Using Transgenic Technology Trait

Effect

Human lysozyme Human lactoferrin Increased κ-casein Increased β-casein Increased α-casein Addition of plasmin inhibitor Addition of β-galactosidase Removal of β-lactoglobulin Addition of desaturase

Increased antimicrobial properties Increased antimicrobial properties Increased heat stability of milk Improved cheese making Improved nutritional qualities Fewer sensory defects of UHT milk Decreased levels of lactose in milk Decreased allergenicity Improved fatty acid profile

Note: Bolded traits have been successfully tested in transgenic animals. Nonbolded traits illustrate potential, but untested applications of transgenic technology. See the text and item 8 in the Recommended Readings List, Mercier and Villotte (1997), for further explanation.

Sheep that have the human gene for α1-antitrypsin (α1AT) are examples of this type of transgenic application. About 100,000 people in the U.S. suffer from α1AT deficiency, a hereditary disease. In this disease, abnormal α1AT genes produce proteins that do not function properly, leading to inflammation and damage to lung cells, which can lead to emphysema. One possible therapy is to give patients normal α1AT; however, sufficient amounts of α1AT cannot be obtained from natural sources. Recombinant α1AT can be obtained from yeast or bacteria, but in both cases it does not have the proper posttranslational glycosylation. It could be produced in mammalian cell lines, but it has proven difficult to achieve sufficient protein yield using cell culture. For these reasons, biotechnologists have attempted to create transgenic animals that produce human α1AT. This was achieved with mice in 1990, and in 1991, scientists created transgenic sheep that produced α1AT in their milk. To date, though, the recombinant proteins produced by these sheep have not been released for human therapy. Transgenic technology also has the potential to modify food-related traits of the milk of cattle and other ruminants. There are a number of objectives to this research (Table 5.3); some are aimed at improving nutritional properties (e.g., increased levels of cysteine-rich α-casein) or decreasing allergenicity (e.g., eliminating or decreasing β-lactoglobulin, which is an allergen, but seems unimportant to the function of milk), whereas others focus on improving processing traits (e.g., increasing the level of β-casein, resulting in milk that forms firmer curds during cheese making). It is much easier to produce transgenic cattle today than it was ten years ago. This is mainly because of improvements in methods of zygote acquisition, maturation of zygotes, and in vitro fertilization of zygotes. As an example, we will examine

144

Introduction to Food Biotechnology

Isolate oocytes from carcasses

Maturation medium

Leutinizing hormone Follicle stimulating hormone

In vitro fertilization

Bull semen

in vitro Injection of desired gene into pronucleus

Growth of embryo to blastocyst stage

Transfer of embryo to cow uterus Gestation in host cow Birth of transgenic cow

FIGURE 5.1 Process used to introduce a desired gene into cattle through pro-injection of pronuclei.

transgenic cows with an introduced human gene coding for α-lactalbumin, the major whey protein in human milk (Figure 5.1). The objective in this case was to produce milk more similar to human milk than conventional cow’s milk. The first step was to obtain large numbers of oocytes (immature, unfertilized egg cells) from freshly slaughtered carcasses. The oocytes were then incubated in a maturation medium, which contains bovine leutinizing hormone and folliclestimulating hormone. These hormones triggered maturation of the oocytes, which is a necessary prelude to fertilization. The oocytes were then fertilized in vitro, using semen collected from a bull. Once the oocyte has been fertilized, it is ready to begin embryonic development. This is a crucial stage in the transgenesis process. Before the fertilized zygote begins to divide, the foreign DNA construct must be injected into one of the pronuclei. This term is given to the nuclei originating from sperm and egg cells that unite to form a diploid nucleus. This is an opportune time to inject foreign DNA, because if DNA is injected at a later stage, after cell division has begun, only part of the embryo will contain the transgene. However, if the fertilized zygote is transformed, all of the cells of the embryo and the adult will contain the transgene. In the α-lactalbumin example, zygotes were incubated after DNA injection in a medium that mimics the environment of the oviduct, where initial growth of the zygote normally occurs. After 7 to 8 days, when the embryos had reached the blastocyst stage (a hollow ball of cells), each embryo was transferred to the uterus

Animal Biotechnology

145

of a cow. After the completion of gestation (about 39 weeks), calves were born and were tested for the presence of the transgene. Nine individuals (five male, four female) were transgenic, and after six months, one of the females was induced to begin lactation. Human α-lactalbumin was present in her milk at a concentration of 2.4 mg/mL. This project illustrates why it is so expensive to produce transgenic cattle. The researchers fertilized 20,918 oocytes; they then injected DNA into 11,507 of the resulting zygotes. This led to 1,011 embryos that were of good enough quality and at the correct stage for implantation in cow uteri. Embryos were implanted in 478 cows — a large herd indeed. This produced 155 successful pregnancies and 90 calves. Of these 90 calves, 9 were transgenic. It is expected that this low success rate is unlikely to improve quickly in the coming years; however, past experience in this field demonstrates that the scope of new achievements is difficult to predict (the surprising announcement of cloning of adult sheep cells in 1997 is a good example).

IV. THE SEARCH FOR EMBRYONIC STEM CELLS It is unfortunate that mice are not suitable for milk or meat production. Embryonic stem (ES) cell lines that can be used to produce germline chimeras have been isolated only from mice, despite numerous attempts to isolate them from agriculturally important animals. What does ES mean and why is it so important? ES cells are isolated from early-stage mouse embryos. With the proper treatment, they can be developmentally arrested; this means that they do not differentiate further and do not complete their normal program of embryogenesis. However, they remain capable of cell division (proliferation), allowing researchers to obtain large numbers of ES cells. The other key feature of ES cells is that after a period of growth in culture, they can be transplanted into another early-stage (blastocyst) embryo. They will then become part of that embryo and remain capable of differentiating into a number of different cell types and tissues (in other words, they are pluripotent). These characteristics are useful because they allow sophisticated manipulation of the ES cell’s genome in vitro. Specific genes can be deleted, added, or modified, and genes can be inserted in specific locations in the genome (targeted insertion). In contrast, pronuclear injection (described in the previous section) allows only the addition of genes, and there is no control over the site of insertion of the new genetic material. When ES cells with altered or additional genes are transplanted into a blastocyst, this gives rise to a chimeric embryo. Some of the cells of the embryo will be unaltered, and some of them will have originated from cell division and differentiation of the ES cells. This embryo will eventually develop into an adult chimera, which has a similar mixture of cells and tissues originating from unaltered or ES cells. This is referred to as a somatic chimera. A germline chimera is a somatic chimera that produces gametes (sperm or oocytes) that are derived from ES cells. This is a crucial characteristic because it allows the use of animal breeding procedures to produce pure breeding lines with the altered characteristic.

146

Introduction to Food Biotechnology Homologous sequences

Thymidine kinase

neo r

Nonfunctional gene

FIGURE 5.2 DNA construct that can be used to make knockout mice that lack a specific gene. (neor = neomycin resistance)

As previously noted, germline chimeras have so far been obtained only from mice. However, somatic chimeras have been obtained from mice, rabbits, and pigs. Many biotechnologists are confident that germline chimeras will eventually be obtained from agricultural animals such as pigs and cattle; if this happens, transgenic food animals will be much easier to produce, and the range of possible genetic modification will be vastly increased. As an example of the use of ES cells to make a transgenic animal, we will examine the procedure used to make a knockout strain, in which a specific gene has been removed from the animal’s genome. The biotechnologist produces a gene construct (Figure 5.2) that has a modified, nonfunctional version of the gene. On both sides of this, there are regions of sequence homology (identical to the sequences on either side of the native gene). A neomycin-resistance gene is also placed between the homologous sequences. Another gene, coding for thymidine kinase, is placed outside of the area between the homologous sequences. A population of ES cells is then electroporated to induce uptake of this construct. As an aside, the process of DNA uptake by mammalian cells is termed transfection rather than transformation, because transformation in the context of mammalian cell culture has traditionally referred to a cellular transition to a cancerous state. A small proportion of cells take up the foreign DNA and insert it into random points in the genome (random insertion) (Figure 5.3). A very small proportion of cells, through the process of homologous recombination, replace the targeted gene with the modified gene present in the construct. It is important to note that only the sequence between the homologous regions in the construct is inserted. Thus, the nonfunctional gene in the construct precisely replaces the native gene. Note also that this means that the thymidine kinase gene will not be inserted into the chromosome. The flanking homologous sequences are essential for homologous recombination. This procedure works because animals, and most other eukaryotes, have DNA repair enzymes that recognize DNA segments that are homologous and splice them together. The cells are then grown in a medium containing neomycin. All cells that lack the integrated construct will then die, because they lack the neomycin-resistant gene. Surviving cells are then transferred to a medium containing the antiviral agent gancyclovir. Cells that have randomly inserted constructs will have the thymidine kinase gene. This enzyme phosphorylates gancyclovir; the resulting compound is a

Animal Biotechnology

147

Isolate embryonic stem cells

Transfect ES cells with construct

gene is * Targeted replaced by the

Construct is inserted at random points in the genome

Nontransfected cells

construct

Incubate cells in medium containing neomycin Nontransfected cells are killed by neomycin Cells survive Cells survive

Incubate cells in medium containing gancyclovir Thymidine kinase phosphorylates gancyclovir

Cells die

Cells survive

Micro-inject into blastocyst

Implant into surrogate mouse

FIGURE 5.3 Process used to make knockout mice through homologous recombination. The pathway leading to knockout mice is bolded and the homologous recombination is indicated by the asterisk (*).

nucleotide analog that will be incorporated into replicating DNA by DNA polymerase. However, it cannot function in transcription, and the cell will die. This will eliminate all cells with randomly inserted constructs. Cells with targeted insertion (i.e., those that result from homologous recombination), though, will lack the thymidine kinase gene, because homologous recombination results in integration of only the DNA between the homologous sequences. The thymidine kinase gene on the construct lies outside of the homologous sequences, so it will not be integrated.

148

Introduction to Food Biotechnology

V. TRANSFER OF SOMATIC NUCLEI: AN ALTERNATIVE TO THE USE OF EMBRYONIC STEM CELLS It is a rare event when a sheep is able to dominate newspaper headlines, as happened in February 1997, when Nature published the first account of cloning of adult somatic cells, in “Dolly.” Wilmut and colleagues did this by isolating oocytes from sheep and then removing their nuclei (this is referred to as enucleation). Nuclei from mammary gland cells of an adult sheep were then transferred into the enucleated cells (this was done by a process similar to protoplast fusion). The cells were then cultured in a medium that promoted embryo development, and the resulting embryos were transplanted into recipient sheep. Viable lambs were born, giving rise to the first mammals cloned from adult somatic cells. This was a revolutionary event because it proved that cell differentiation in mammals is not irreversible. A mammary cell nucleus, when placed in the cytoplasmic environment of an unfertilized oocyte, was able to dedifferentiate and form an embryo with its hundreds of different cell types; it also was able to form a functioning embryo that had all these cell types in the right place at the right time. Adult cell cloning is a technique that could potentially be used to clone human beings; consequently, it is highly controversial, and much debate has ensued on how to prevent such use of the technology. However, adult cell cloning has many applications that are less controversial. For example, similar techniques can theoretically be exploited to use primary cultures of mammalian cells as the source of genomic DNA for a developing embryo. In this case, the oocytes would be enucleated and replaced with a nucleus from a primary culture. Primary cultures can be obtained from many adult organs, and they can usually be cultured for numerous passages (moving from one flask to another). This means that gene targeting is possible, because the procedures of homologous recombination, followed by selection of transfected cells that are used with ES cells, can also be used with primary cultures. Nuclei from successfully transfected cells can then be transferred to enucleated oocytes, and embryos can then be induced from the oocytes. In June 2000, another paper appeared in Nature that described successful gene targeting using this strategy. The human α-antitrypsin gene was inserted into the sheep genome at the end of a procollagen gene (Figure 5.4). This particular procollagen gene was selected because it is well characterized in sheep and is constitutively expressed in fibroblast cells (these cells were desirable for a number of reasons). The α-antitrypsin transgene was then expressed as a fusion protein linked to the procollagen protein. Fusion proteins result from translation of mRNA transcribed by adjacent genes controlled by the same promoter. Consequently, they are a composite of two proteins. Transfected cells were introduced into enucleated oocytes, and the resulting embryos were transferred to recipient ewes. Three viable lambs were born; after 1 year, one lamb was induced to lactate, and the milk contained human α-antitrypsin at a concentration of 650 µg/mL. This technology is not perfect; both cloned and transgenic sheep suffer from several abnormalities, including a tendency to obesity. A large proportion of embryos derived in this way die during gestation, often because of abnormal kidney or brain

Animal Biotechnology

149

Design construct with human α-antitrypsin gene

Transfect sheep fibroblasts Homologous recombination Isolate oocytes from sheep

procollagen gene is replaced by α-antitrypsin gene

Isolate nuclei

remove oocyte nucleus (enucleation)

Introduce recombinant nucleus into enucleated oocyte

Oocyte develops into embryo

Implant into surrogate sheep

FIGURE 5.4 Introduction of the human α-antitrypsin gene into sheep using homologous recombination and enucleated oocytes.

development. These problems must be solved before the technology can be widely applied to the development of transgenic animals with altered properties for food and agriculture. Nevertheless, the process of adult cell cloning has given biotechnologists an important option in projects aimed at improvement of food animals.

RECOMMENDED READING 1. Sang, H., Transgenic chickens — methods and potential applications, Trends Biotechnol., 12, 415, 1994. 2. Sheldon, B. L., Research and development in 2000: directions and priorities for the world’s poultry science community, Poultry Sci., 79, 147, 2000. 3. Devlin, R. H., Transgenic salmonids, in Transgenic Animals: Generation and Use, Houdebine, L. M., Ed., Harwood Academic Publishers, Amsterdam, 1997, chap. 19.

150

Introduction to Food Biotechnology

4. Dunham, R. A. and Devlin, R. H., Comparison of traditional breeding and transgenesis in farmed fish with implications for growth enhancement and fitness, in Transgenic Animals in Agriculture, Murray, J. D., Anderson, G. B., Oberbauer, A. M., and McGloughlin, M. M., Eds., CAB International, Wallingford, U.K., 1999, chap. 6. 5. Muir, W. M. and Howard, R. D., Possible ecological risks of transgenic organism release when transgenes affect mating success: sexual selection and the Trojan gene hypothesis, Proc. Natl. Acad. Sci. U.S.A., 96, 13853, 1999. 6. O’Flynn, F. M., McGeachy, S. A., Friars, G. W., Benfey, T. J., and Bailey, J. K., Comparisons of cultured triploid and diploid Atlantic salmon (Salmo salar L.), ICES J. Mar. Sci., 54, 1160, 1997. 7. Devlin, R. H., Risk assessment of genetically-distinct salmonids: difficulties in ecological risk assessment of transgenic and domesticated fish, in Aquaculture and the Protection of Wild Salmon, Gallaugher, P., Ed., Simon Fraser University, Vancouver, 2000. 8. Mercier, J.-C. and Vilotte, J.-L., The modification of milk protein composition through transgenesis: progress and problems, in Transgenic Animals: Generation and Use, Houdebine, L. M., Ed., Harwood Academic Publishers, Amsterdam, 1997, chap. 70. 9. Murray, J. D. and Maga, E. A., Changing the composition and properties of milk, in Transgenic Animals in Agriculture, Murray, J. D., Anderson, G. B., Oberbauer, A. M., and McGloughlin, M. M., Eds., CAB International, Wallingford, U.K., 1999, chap. 14. 10. Eyestone, W. H., Production of transgenic cattle expressing a recombinant protein in milk, in Transgenic Animals in Agriculture, Murray, J. D., Anderson, G. B., Oberbauer, A. M., and McGloughlin, M. M., Eds., CAB International, Wallingford, U.K., 1999, chap. 13. 11. Anderson, G. B., Embryonic stem cells in agricultural species, in Transgenic Animals in Agriculture, Murray, J. D., Anderson, G. B., Oberbauer, A. M., and McGloughlin, M. M., Eds., CAB International, Wallingford, U.K., 1999, chap. 4. 12. Wilmut, I., Schnieke, A. E., McWhir, J., Kind, A. J., and Campbell, K. H. S., Viable offspring derived from fetal and adult mammalian cells, Nature, 385, 810, 1997. 13. McCreath, K. J., Howcroft, J., Campbell, K. H. S., Colman, A., Schnieke, A. E., and Kind, A. J., Production of gene-targeted sheep by nuclear transfer from cultured somatic cells, Nature, 405, 1066, 2000.

6

Diagnostic Systems I. WHY ARE DIAGNOSTIC SYSTEMS NEEDED?

A. OVERVIEW

AND

GLOBAL PERSPECTIVE

It is impossible to completely eliminate pathogens from the food supply; some pathogens (e.g., Bacillus cereus) are common in soil and on vegetation, and food handlers often carry others (e.g., Staphylococcus aureus), even if they follow standard hygiene practices. Lapses in worker hygiene or sanitation in food processing plants results in a dramatic expansion of the range of potential pathogens, because of the broad distribution of pathogenic bacteria and viruses. This ubiquity of pathogens makes it essential that the food industry have access to efficient diagnostic tools that allow detection and identification of pathogens. Diagnostic tools are also essential for clinicians, to help in diagnosis of foodborne illness, which is usually classified as gastroenteritis (inflammation of the stomach or, small or large intestines). Many pathogens cause similar symptoms, including severe abdominal pain, diarrhea (watery stools), and vomiting. Identifying the culprit is important because effective treatments vary among the pathogens; for example, antibiotic treatment of Escherichia coli O157:H7 is not usually effective, whereas antibiotic treatment of Clostridium difficile infections is effective and necessary. Accurate identification of the cause of food-borne illness is also useful to public health officials attempting to identify the source of an outbreak, and for epidemiologists interested in long-term trends, such as the increasing frequency of E. coli O157:H7 infections. Public health officials identify sources by testing the stools of affected people for the presence of pathogens. Once a pathogen has been identified, investigators carefully interview victims and attempt to uncover commonalities. For example, all the victims may have eaten at the same restaurant the previous day or may have attended a recent family reunion. In some cases it is essential to identify all the people affected by an outbreak, so that they can be closely monitored for the eruption of severe symptoms. For example, children under the age of 5 are at risk for developing potentially life-threatening kidney failure after infection by E. coli O157:H7; it is essential to identify this pathogen as soon as possible. Diagnostic speed is also essential in food processing plants. If pathogen contamination can be detected before a product lot leaves the plant, it may be possible to avert an outbreak. However, if enough time elapses before contamination is detected, it may be too late to prevent an outbreak, depending on the nature of the product and its distribution system. Current methods of microbial identification often require 2 to 3 d, which increases the risk that contamination of food by pathogens will be undetected until it is too late.

151

152

Introduction to Food Biotechnology

TABLE 6.1 Emerging Pathogens and Examples of Outbreaks Pathogen

Outbreak Location

Year

Food Implicated

E. coli O157:H7 Cyclospora Cryptosporidium Listeria monocytogenes Calicivirus

Osaka, Japan U.S. & Canada U.S. Nova Scotia, Canada Cruise ship

1996 1996 1996 1981 1993

Radish Sprouts Raspberries Apple cider Sauerkraut Fresh-cut fruit

Number Affected >8000 1465 160 41 217

Food-borne pathogens and difficulties in detection and identification are a continuing worry for the food industry and public health authorities, partly because the incidence of emerging pathogens (Table 6.1) is increasing. E. coli O157:H7, for example, was first implicated as a food-borne pathogen in 1982. Since then, it has caused a number of large outbreaks throughout the world. Because the principal reservoir of this bacterium is in cattle, outbreaks are usually linked to undercooked beef or to contamination of food or water by cattle manure. This pathogen causes a wide range of severity of symptoms, but it is most serious in the elderly, in whom it often leads to severe dysentery, and in the very young, in whom it may cause kidney failure. E. coli O157:H7 poses a challenge to developers of diagnostic tests because the infective dose (number of bacteria required to cause illness) is extremely small (as low as 50 organisms). Fortunately, it can be differentiated from nonpathogenic strains by its inability to ferment sorbitol. To make matters more confusing, a number of other strains of E. coli cause similar illnesses as E. coli O157:H7 but are able to ferment sorbitol. Efficient identification of these strains is difficult with current technologies. Cyclospora is another example of an emerging pathogen. This protozoan is similar to Cryptosporidium, which has a much longer history as a cause of foodand water-borne illness. Both are protozoans belonging to the same group (apicomplexans) as Plasmodium, the cause of malaria. In 1996, Cyclospora-contaminated raspberries imported from Guatemala led to 1465 cases of cyclosporiasis, which is characterized by diarrhea that lasts from 1 to 6 weeks. It is unclear how the raspberries became contaminated, but the most likely explanation is that water used to mix insecticide and fungicide formulations was contaminated with oocytes (dormant egg-like structures produced by apicomplexans) of Cyclospora. This organism is difficult to detect in clinical (stool) or environmental (soil or water) samples — unlike most bacteria, it cannot be grown on agar media. It can be detected only by microscopic examination of stained samples. The development of more sensitive diagnostic tests would increase our power to monitor the distribution of this parasite and understand how it contaminates food. Another reason for concern about food-borne pathogens is that the incidence of food-borne illness is increasing worldwide. This is partly because of increased reporting of gastroenteritis and acknowledgment that many cases formerly ascribed to “stomach flus” are in fact caused by food-borne viruses such as the Norwalk virus

Diagnostic Systems

153

(a calicivirus). But the true incidence also seems to be increasing, particularly in industrialized countries. One reason is that consumers demand access to fresh produce year round, which leads to large-scale importation of fruits and vegetables (cyclospora in raspberries is an example of this). Food-borne pathogens are also a major concern in the developing world. Diarrhea caused by pathogenic microbes and parasites is estimated to lead to more than 2 million deaths yearly in the developing world. Solving this immense problem requires development and distribution of inexpensive water purification and sewage treatment systems. In the developed world, water purification is achieved through centralized plants that filter particles, absorb chemicals, and chlorinate the water before sending it through a system of pipes to homes and businesses. Unfortunately, this type of water purification is prohibitively expensive for many communities in the developing world; some researchers recommend that interim, inexpensive systems (e.g., chlorination by homeowners) be implemented to treat water until community-level systems can be built. Also, the World Health Organization (WHO) is currently setting up improved systems for reporting of food- and water-borne diseases and outbreaks, as well as databases focusing on the distribution of the important food-borne pathogens throughout the developing world. The development of cheap, accurate diagnostic systems would greatly aid this important goal.

B. DIAGNOSTICS

AND

HAZARD ANALYSIS CRITICAL CONTROL POINTS

Food companies have traditionally relied on end product inspection and testing to ensure that the food that leaves the factory does not contain an excessive load of nonpathogenic and pathogenic microbes. Theoretically, 100% of products can be visually inspected, but human frailties (e.g., distractability, boredom) decrease the efficiency of inspection. Furthermore, many microbial defects cannot be detected by visual inspection. Hence, destructive sampling of end products is often required. In principle, this is a simple process — a defined proportion (subset) of products is removed at the end of the production line and the level of microbial contamination is assessed. Typically, microbes are put into suspension by grinding each sample in a Stomacher™ or similar apparatus. A dilution series of this suspension is then plated onto an agar medium, and after a suitable period (24 h at 37°C), bacterial or fungal colonies are counted. In some cases, enrichment and selective media are used to detect specific pathogens (e.g., Salmonella) (Figure 6.1). This technology can be quite powerful; in some cases, cultural methods can distinguish closely related strains. For example, E. coli O157:H7 can be distinguished from nonpathogenic strains through the use of selective and differential media (differential media produce a detectable change in appearance/color of either the agar medium or the colonies when the target bacterium is present). Microbiologists have also devised clever techniques to increase the speed and accuracy of culture-based diagnostics. For example, de Boer and colleagues in the Netherlands developed a medium for the detection of Salmonella that is selective and differential and makes use of Salmonella’s motility to help identification. Salmonella are visible as a migrating pink zone within a green medium. This medium allows identification of Salmonella 1 d earlier than usually possible.

154

Introduction to Food Biotechnology Put food sample in stomacher Resuscitate Salmonella 24 hours Enrich for Salmonella 24 hours Select Salmonella 24 hours Confirm using biochemical tests or serology

FIGURE 6.1 Detection of Salmonella in food using cultural methods.

Cultural methods such as this are a reliable indicator of microbiological quality in food, but there are many disadvantages of end product microbial analysis. Financial concerns sometimes make it difficult to test an adequate number of samples, because cultural methods are labor and materials intensive. If too few samples are assessed, the risk of missing contaminated product increases. This risk also increases if contamination occurs sporadically, rather than on a regular basis. Also, cultural methods are slow, particularly if the aim is to identify specific pathogens. For example, conventional detection of Salmonella (see Figure 6.1) requires that a food sample be incubated in three successive media for a total time of 72 h. Additional tests that confirm the presence of Salmonella are then required. This confirmation can be biochemical; the colonies that are presumed to be Salmonella are typically inoculated into a series of media containing a range of different carbon sources (e.g., glucose, sucrose, mannitol, proline, etc.). The bacteria respiring or fermenting (depending on the diagnostic system) these carbon sources produce a “pattern of utilization” that can be used to identify the bacteria. Unfortunately, the growth period required for these tests further adds to the delay in assessment of contamination. Antibody-based tests, in contrast, can confirm the presence of Salmonella immediately. However, if negative, they do not provide any other information that could be used for identification, unlike biochemical tests. Resuscitation and selective media for cultural methods of identification are necessary because it is very difficult to detect small numbers of specific microbial species, especially if there is a large background of nonpathogenic bacteria and fungi. A food that has just been cooked at a high temperature is unlikely to have a large background, but many other foods (e.g., yogurt and cheese) typically harbor large populations of useful or harmless bacteria and fungi. Because of the time, labor, and financial constraints of end product testing, and because of its intrinsic inefficiency (in most cases, it is impossible to take enough samples to achieve a satisfactory level of protection), most government regulators

Diagnostic Systems

155

Process Steps

Hazards

Receive beef

Temperature abuse leads to microbial growth

Store beef

Temperature abuse, cross contamination

Cook

Insufficient cooking leads to survival of pathogens

Assemble hamburger (add bun, tomatoes, etc.)

Cross contamination (e.g. raw hamburger drips onto tomatoes)

Hold till served to consumer

Temperature abuse leads to microbial growth

FIGURE 6.2 Hazards associated with each step in the processing of hamburgers at a fastfood restaurant. For each step, a strategy would be devised for control of hazards and for monitoring. The aim of monitoring is to ensure that control is achieved on a day-to-day basis.

and standardization organizations (e.g., International Organization for Standardization [ISO]) agree that hazard analysis critical control points (HACCP) systems can increase food safety in production lines. HACCP processes aim to prevent contamination of food by pathogens, rather than simply detect contamination after it has happened. HACCP analysis begins with a thorough understanding of each step in the process. Next, the hazards associated with each step are identified (Figure 6.2). Although devised for industrial food processing plants, HACCP principles can also be applied to domestic, restaurant, catering, or even agricultural processes. Common hazards include contaminated raw materials (e.g., Salmonella and Campylobacter in raw chicken), cross contamination (e.g., contamination of vegetables by contact with raw meat), or temperature abuse that allows microbial growth. For example, if foods are not refrigerated during holding times, bacteria such as S. aureus or B. cereus could grow and produce toxins. The HACCP plan then outlines steps that can be taken to control these hazards. In each process, certain steps are crucial, in terms of food safety. They are designated as critical control points (CCPs). CCPs often consist of heat treatments that eliminate or reduce the number of microbes. Inadequate performance of a CCP (e.g., heating at too low a temperature) could lead to potential contamination of the food and potential transmission of pathogens to the consumer. Once the hazards and CCPs in a process have been identified, they must be monitored. This can often be done by checking the operating parameters (e.g., temperature) of processing machinery. It is often useful to monitor control points

156

Introduction to Food Biotechnology

using microbial assays, aimed at detection of specific pathogens, or assessment of overall numbers of contaminating microbes (microbial load). However, time and labor constraints associated with traditional cultural methods may make it difficult to effectively monitor the efficacy of a HACCP system. For example, HACCP plans for a poultry processor could include periodic tests for the presence of Campylobacter in chicken parts. This is statistically feasible in most regions because of the high frequency of Campylobacter contamination in chickens entering a processing plant. However, primarily because of the time involved in testing for Campylobacter, monitoring is usually based on total plate counts, which estimate the total numbers of contaminating microbes. Unfortunately, total plate counts are unlikely to be an accurate indicator of Campylobacter contamination. However, if rapid diagnostic tests were available, they could be used to effectively monitor levels of Campylobacter and other pathogens. This is one example of the need for rapid diagnostic tests in the food industry. Ideally, detection of pathogens should take less than 24 h, which allows food processors to take corrective action if contamination occurs. Some biotechnology companies are also attempting to develop on-line methods of microbial monitoring that could be used to continuously monitor microbial parameters; these will be discussed later in this chapter, in the context of biosensors. HACCP programs should always undergo a verification procedure. Diagnostic methods are useful here because they can tell HACCP developers if the program is successful in preventing food contamination by pathogens. Regulatory authorities sometimes require HACCP verification as part of the certification process.

C. NONPATHOGEN DIAGNOSTICS Most commercial diagnostic tests are aimed at the detection of pathogens, but a number of tests target nonpathogenic microbes, especially spoilage agents (Table 6.2). Food processors often incur heavy losses through microbial spoilage. Many of the topics discussed above in the context of pathogens (outbreak investigation, HACCP) are also applicable to food spoilage. For example, if a large lot of prunes is spoiled by growth of a xerophilic (dessication-loving) fungus, the company involved needs to quickly identify the fungus responsible and the factors that allowed it to contaminate and grow in the product. Then, a HACCP plan can be developed to prevent further incidents. Mycotoxin contamination is also a great worry to the food industry. Some of the fungi that commonly contaminate food crops (e.g., Aspergillus flavus and Fusarium spp.) produce potent mycotoxins (aflatoxins and tricothecenes, respectively) that are a significant public health hazard. Consequently, governments throughout the world have established allowable limits for several mycotoxins in food. For example, the U.S. Food and Drug Administration (FDA) allows 20 µg/kg (ppb) contamination of aflatoxins in food (this is primarily aimed at nuts and nut products). Allowable levels for animal feed are substantially higher (e.g., 300 ppb in animal feed for cattle bound for slaughter). Due to a lack of universal agreement on safe levels of aflatoxins, allowable limits vary among countries. In Canada, for example, a limit is set for nuts destined for human consumption (15 ppb) and for all animal feeds (20 ppb). In the European Union (EU), draft regulations call for extremely low limits (6 ppb)

Diagnostic Systems

157

TABLE 6.2 Diagnostic Applications Relevant to Food Production and Processing Goal of Diagnostic

Upcoming Developments

Example

Current Technology

Assess microbial contamination

Bacteria on carcasses

Plate counts

Identify pathogen

E. coli O157:H7

Identify spoilage agent

Yeasts

Investigation outbreak

Tracing pathogen strains to source

Selective media, immuno-assay, biochemical, DNA tests Biochemical identification, immuno-assay Biochemical tests, immuno-assay

Monitor microbial growth

Yeast fermentation

Plate counts

Monitor hygiene

Work surfaces in processing plant

Plate counts/ATP using bioluminescence

Process monitoring

Glucose consumption in bioreactor Aflatoxin

Analytical chemistry, biosensors Immuno-assay Immuno-assay

Gluten

Analytical chemistry

Antibody based

Pork in all-beef products Herbicide-resistant soybeans

Detection of unwanted protein DNA test or immunoassay

Detection of unwanted DNA Improved sensitivity and reliability of DNA-based tests

Detect toxin residues Detect pesticide residues Detect residues linked to allergy/intolerance Detect adulteration Detect transgenic crops in food a b c

Flow cytometry, impedimetry, biosensors Bioluminescence, biosensors

DNA tests

DNA tests (RAPD,a RFLP,b pulsed-field electrophoresis, 16S rRNAc sequence analysis) Biosensors, flow injection, flow cytometry Increased specificity (combining hygiene tests with pathogen identification) More biosensors Biosensors

RAPD, randomly amplified polymorphic DNA. RFLP, restriction fragment length polymorphism. rRNA, ribosomal RNA.

of aflaxotins in nuts for human consumption. Evidently, we need efficient diagnostic tests to monitor levels of aflatoxins and other mycotoxins, especially considering the scope of international trade in commodities (e.g., wheat, peanuts) that are susceptible to mycotoxin contamination. Currently, most mycotoxin detection is done

158

Introduction to Food Biotechnology

through antibody-based tests or analytical chemical techniques (e.g., high-performance liquid chromatography [HPLC]). Food adulteration is an ongoing problem in the food industry, particularly in processed meats. For example, unscrupulous companies incorporate pork or other meats into products labeled “all beef.” Diagnostic tests that detect such adulteration are commercially available, and are usually based on the ability of antibodies to differentiate between muscle proteins from different species. Sometimes, processors inadvertently allow contamination of food by unwanted ingredients. For example, a food that is labeled “gluten-free” may become contaminated by gluten-containing wheat introduced through one of the food ingredients. Because the major market for gluten-free products is people who have gluten intolerance, the presence of contaminating wheat is clearly “intolerable.” Diagnostic tests have the potential to reduce the risk of this happening. In 1999, polymerase chain reaction (PCR)-based diagnostic tests were used to detect Starlink corn in food. This cultivar of transgenic corn had been approved for feed use by the FDA, but not for food. The ensuing controversy, which reverberates to this date, demonstrated the difficulties associated with segregation of transgenic crops, as well as the urgent need for effective and cheap diagnostic tests for transgenic crops. Reliable tests to detect transgenic crops will increase in importance in the future, as the EU and other countries decide on allowable limits of transgenic crops in food. Current proposals call for a 1% limit for transgenic residues in food. If the level is greater than 1%, the food label will be required to indicate that the food contains transgenic (genetically modified) organisms. Finally, diagnostic techniques are used in breweries, cheese factories, and dairy plants to identify useful microbes. These processes use specific strains of yeasts and lactic acid bacteria, and the presence of the wrong strain can adversely affect product quality. This is a challenging problem for developers of diagnostic methods, because many strains of lactic acid bacteria are physiologically and genetically very similar, making it difficult to differentiate among them using cultural methods. Yeast strains used in the brewing and wine-making industries are similarly difficult to differentiate. DNA-based methods are increasingly used in this context.

II. DIAGNOSTIC BIOTECHNOLOGY A. SCOPE Biotechnology can dramatically reduce the time and labor required to detect and identify microbes. The following applications of biotechnology to diagnostics are discussed in this chapter. • Gene probes to detect pathogenic microbes • PCR to detect contaminating pathogens or the presence of transgenic crops • DNA chips and micro-arrays • Antibodies in diagnostic systems

Diagnostic Systems

159

• Bioluminescence to monitor hygiene and contamination by specific microbes • Biosensors in the food industry Antibody-based assays that facilitate detection are available for most major pathogens and are particularly important for identification of food-borne viruses. Nucleic-acid-based kits for the detection of Salmonella, Campylobacter, E. coli, S. aureus, and Listeria monocytogenes are also commercially available. Nucleicacid-based techniques are also becoming increasingly important in the typing (strain identification) of bacteria.

B. NUCLEIC ACID PROBES DNA and RNA probes can be highly specific; they can be designed so that they hybridize only to the target species or strain. All that is required is identification of a DNA sequence that is unique to the target microbe. The use of nucleic acid probes is well established in molecular biology (see Chapter 2, Section V.H.3). Biotechnologists have long recognized the potential of probes in diagnostic systems, and this potential is now starting to be realized. However, we needed to transform traditional hybridization procedures such as southern blotting, which are labor intensive and “fussy,” into more user-friendly and safer techniques. For example, when DNA probes were first used in molecular biology, they were labeled with radioactive isotopes (e.g., 32P). This works well in a research laboratory, where the problems associated with the use of radioisotopes (occupational safety, disposal of wastes) can be safely and adequately addressed. However, few quality control laboratories in food companies have the necessary equipment for radioisotope work. The requisite training of personnel in the use of radioactive compounds is an additional problem for food companies. For these reasons, nonradioactive detection systems are used in commercial versions of probe-based assays. For a diagnostic test to achieve widespread use in the food industry, it must also be easy to use. Probe-based tests have achieved this goal through immobilizing probes to inorganic supports (dipsticks) that allow the user to easily manipulate the probe (e.g., wash off unhybridized DNA) without damaging or losing it. This is referred to as solid-phase hybridization; other hybridization systems (e.g., solutionbased) are also possible. The principle behind the use of DNA probes is quite simple (see Figure 2.15). Short single strands of DNA that are complementary to genes present in a pathogenic microbe are synthesized. The food sample must then be treated so that any microbial cells are lysed, releasing their DNA. Microbial DNA is then treated to convert it from double strands to single strands, and the probe is added. Hybridization (annealing of complementary strands) then occurs between the single-stranded DNA probe and single-stranded DNA released from pathogenic microbes present in the food. Probe DNA that has not hybridized must then be removed, usually by washing the sample, and the presence of hybridized DNA probes can then be detected. To understand how the above process works, we can examine the Gene Trak™ system for detection of Salmonella in food samples (Figure 6.3). This system uses

160

Introduction to Food Biotechnology

Stomach food sample

Salmonella cells

Lyse cells and convert ds DNA to ss DNA

AAAAAA

Add capture probe

* AAAAAA

and detector probe

Salmonella DNA

TTTTTT

AAAAAA

*

transfer dipstick

Salmonella DNA

*

Solution containing enzyme substrate Look for color change

FIGURE 6.3 The use of a gene probe to detect a pathogen in food. The Gene Trak™ system for the detection of Salmonella is used as an example. The capture and detection probes anneal to different regions of the ribosomal DNA gene of Salmonella. The dipstick is used to remove the capture probe–Salmonella DNA–detection probe complex. The complex is then placed in the appropriate solution that will reveal the presence of the detection probe. In earlier versions of this system, detection was based on fluorescein in the probe binding to antifluorescein antibodies, which in turn bind to enzyme-linked antibodies. The enzyme then catalyzes formation of a colored end product.

a capture probe and a detection probe. These probes hybridize to different regions of the genes coding for ribosomal RNA (rRNA) in Salmonella. rRNA genes were selected because DNA probes work best if the target DNA is highly copied (i.e., many copies of the gene are present in the cell). Around 5000 copies of rRNA genes are present in Salmonella. The large amount of rRNA in a bacterial cell is also important because most of the binding between probe and target is through rRNA binding to the probe. It is also essential that the DNA probes be specific to Salmonella;

Diagnostic Systems

161

they must not bind to related bacterial species such as E. coli. Specific rRNA sequences were found by comparing the rRNA sequences of a large range of bacteria and selecting sequences that are specific to Salmonella. So, how does Gene Trak work? Enrichment of Salmonella is required, in order to produce detectable amounts of Salmonella DNA. However, time savings are achieved after enrichment, because time-consuming growth in selective media is not required. Thus, Gene Trak can detect Salmonella after 48 h (sometimes after 24 h), at least 1 d sooner than through conventional methods. After enrichment, bacteria in the food sample are lysed using NaOH, and the capture and reporter probes are added. Note that the capture probe contains a poly A sequence — a sequence of adenine residues. Note also that the reporter probe is covalently bound to fluorescein. Both probes hybridize to complementary regions of DNA released from Salmonella cells. The next step requires the removal of unbound capture and reporter probes. This is done by adding a dipstick that contains strands of poly T (a sequence of thymine residues). The poly T strands hybridize to the poly A tails on the capture probe. The crucial point is that if the capture probe has hybridized to a strand of Salmonella DNA, then that strand of Salmonella DNA has probably also hybridized to the reporter probe. Therefore, when the dipstick is removed from the solution, it will carry with it: (1) unhybridized capture probes and (2) strands of Salmonella DNA that have hybridized to both capture and reporter probes. The presence of unhybridized capture probe on the dipstick is irrelevant, because detection of the captured Salmonella strand is based on the presence of fluorescein in the reporter probe. Unhybridized capture probes will not contain fluorescein and will not be detected. The final step involves detection of fluorescein. Several alternatives are possible; the Gene Trak method uses an antibody-based system. Antibodies that bind specifically to fluorescein are added. The antibodies are covalently linked to an enzyme (horseradish peroxidase) that can convert a synthetic substrate (chromogen) into a colored end product. Thus, Salmonella is detected through a change in color of the solution. The principle advantage of this system is the reduced time required for positive identification of Salmonella. Another advantage is the assay’s specificity; several bacteria found in food (Citrobacter, Enterobacter, Escherichia, Klebsiella, and Proteus) are very closely related to Salmonella and are often able to grow in media that are “selective” for Salmonella. Furthermore, some isolates of Salmonella are atypical and do not produce the usual colony characteristics of Salmonella on differential media. The main problem with the Gene Trak system is that enrichment of Salmonella is still required, making “instant” identification of Salmonella impossible. However, enrichment is in a sense useful, because it makes it unlikely that dead Salmonella will be detected, unless they are present in high numbers. This is important, because dead Salmonella do not pose a health hazard, and the detection of dead Salmonella constitutes a false positive that may be wasteful of a company’s resources. Because it is still possible that dead bacteria could lead to a false postitive, positive probe results are usually confirmed through traditional cultural methods. Similar probe-based tests are available from at least one other company (Accuprobe®) for a range of clinically and food-relevant pathogens.

162

C. EXPLOITING

Introduction to Food Biotechnology THE

POLYMERASE CHAIN REACTION

PCR has the potential to allow extremely rapid (several hours) identification of pathogenic microbes. As explained in Chapter 3, PCR can quickly amplify specific sequences of DNA and can theoretically amplify a single copy of a DNA sequence of a sample. Thus, PCR has the potential to realize a long-sought goal of food diagnostics: the detection of a single pathogenic microbe in a 25-g food sample within a few hours. Unfortunately, the use of PCR in food samples is technically challenging; food usually contains a large background of plant or animal DNA that might interfere with the PCR, as well as chemicals that inhibit the PCR, sometimes by directly inhibiting DNA polymerase. These problems have been addressed for certain foods, through dilution of inhibitors or concentration of bacterial cells. A number of strategies are possible for concentration, including centrifugation of liquid samples or liquefaction of solid foods, followed by an affinity separation technique. One affinity technique involves covalently binding pathogen-specific antibodies to magnetic beads. The magnetic beads are then added to a suspension of a food sample, and, after incubation, the beads (and any beads bound to pathogens) are separated magnetically. The presence of low numbers of pathogens can then be revealed through PCR. Magnetic beads have also been used to increase the efficiency of E. coli O157:H7 detection using cultural methods. PCR-based diagnostic kits for detection of food-borne pathogens (e.g., L. monocytogenes) are now commercially available. One company (Biotechon) has developed a system that allows detection of amplified products during amplification. This is done through the use of fluorescent probes that are added to the PCR mixture. One probe is a “donor” compound and another probe is an “acceptor.” The donor compound absorbs light energy of a specific wavelength and transfers it via resonance energy to the acceptor compound. This results in emission of light at a higher wavelength than the original light, which is detected electronically. This fluorescence happens only when the donor and acceptor compound are in close proximity, as when both probes are hybridized to amplified DNA specific to the target bacterium. It is too early to tell if the food industry will embrace PCR-based diagnostics. Considerable expenses are involved in the acquisition of thermal cyclers and the training of personnel in their use. Also, these techniques need to be extensively validated, through comparison of their ability to detect low levels of contamination in food to those of conventional culture-based methods. One recently published study (Bellin et al., 2001) demonstrated successful use of LightCycler™ to detect a food-borne pathogen (strains of E. coli, including O157:H7, that produce Shiga toxins). However, this study detected the bacterium from pure cultures, not from food. The ability of PCR to detect dead organisms is a problem. Tiny amounts of DNA released from pathogens killed by heat treatments, for example, would be amplified by PCR. This is less of a problem if enrichment precedes PCR, but that extends the procedure to at least 1 d. In some situations the presence of live or dead pathogens is an important indicator of food safety. For example, if PCR detects Clostridium botulinum in a food sample, it is a cause for concern whether the bacterium is alive or dead, because dead bacteria may have produced botulinum

Diagnostic Systems

163

toxin before dying. The toxin could then persist in the food, causing a potential for serious illness. One final note on PCR: several alternative amplification systems are actively under development. For example, nucleic acid sequence based amplification (NASBA®) uses three viral enzymes to amplify either RNA or DNA targets. The main advantage of this technique is that it is isothermal (occurs at a constant temperature), avoiding the need for expensive thermal cyclers. So far, NASBA diagnostics have mainly been aimed at identification of viruses, but applications for the identification of Campylobacter and L. monocytogenes are also under development.

D. DNA CHIPS

AND

MICRO-ARRAYS

Few techniques have caused as much excitement among microbiologists as DNA chips and micro-arrays. Both are intrinsically miniaturized extensions of conventional tests of nucleic acid hybridization. Imagine a membrane that is used for a dot blot (Chapter 2, Section V.H.3). A sample of DNA is placed onto the membrane; labeled probe is then added; and hybridization (if present) is detected. Now consider a slightly different scenario: a number of oligonucleotides (each representing a different gene) are immobilized onto separate points of a membrane. A bacterial culture is then exposed to a chemical that results in labeled messenger RNA (mRNA) transcripts. The bacteria are lysed and then placed on each oligonucleotide on the membrane. After washing, hybridization between labeled mRNA and immobilized oligonucleotides can be detected. This macro-array of oligonucleotide probes allows simultaneous detection of expression of a number of genes. Now imagine a similar array of immobilized oligonucleotides on a 1 × 1 cm square on a glass microscope slide. Further miniaturization can be achieved with tiny wells etched into a circuit board. Oligonucleotides are then immobilized onto these wells, and the pattern of probe immobilization (i.e., which probes are loaded into which wells) can be controlled electronically. These are often referred to as DNA chips or micro-arrays (arrays on glass slides are also commonly referred to as micro-arrays). The great advantage of these systems is that they require only small amounts of resources. DNA chips can also be developed into laboratories in a chip, wherein an experimental routine, perhaps involving heating of reagents and mixing of several different chemicals, or even electrophoresis, can occur at a micro scale. If successfully applied to diagnostic testing, micro-arrays embedded into silicon chips could allow efficient testing of large numbers of samples and could even be used to amplify sample DNA using PCR. Theoretically, with one test the investigator could detect the presence of DNA of many different pathogens in a food sample. Micro-arrays are currently commercially available for assessment of global gene expression (the total pattern of gene expression by a cell). Up to 8000 genes can be simultaneously tested for hybridization, allowing investigators to dissect changes in gene expression after different experimental treatments. This type of research has many potential applications in food microbiology (e.g., determining pattern of gene expression in a pathogenic bacterium triggered by exposure to acidic preservatives). Micro arrays also have great potential as diagnostic systems, but this is currently in the research and development phase.

164

Introduction to Food Biotechnology

E. ANTIBODY-BASED DIAGNOSTIC SYSTEMS 1. Applications of Antibodies to Diagnostics Antibodies are proteins produced by the mammalian immune system. Their biological function relies on their ability to bind specifically to proteins and other compounds. Antibodies are produced by B-cells when foreign compounds invade the body of a mammal. For example, the presence of S. aureus in the blood will trigger the production of antibodies that bind specifically to proteins on the surface of S. aureus cells. Live microbes are not essential for this process; microbial products (e.g., exotoxins) or components (e.g., cell wall fragments) also provoke the production of specific antibodies. There are five classes of antibodies — IgA, IgD, IgG, IgM, and IgE. IgG is the most frequently used in diagnostic tests; it is a y-shaped protein made up of a constant region (the stalk of the y), which is the same in all IgG molecules, and a variable region (the top of the y), which is a mixture of regions that are similar among different molecules and regions that are specific to each molecule. Thus the variable regions give antibodies their specificity of recognition. The molecule (usually a protein) that is recognized by an antibody is an antigen. Antibodies exhibit variable levels of specificity (ability to bind to one antigen and not to others) and affinity (a measure of the strength of antigen–antibody binding). In a diagnostic test, it is important that the antibodies that are used have high specificity (thus avoiding cross reactions with similar antigens) and high affinity, which allows the use of low concentrations of the antibody. Cross reactions can lead to expensive false-positive reactions; for example, if an antibody is used in a diagnostic test to detect Vibrio vulnificus, it must not bind to other, nonpathogenic species within the Vibrio genus. Antigens have numerous epitopes representing different regions of antibody specificity. When a pathogen invades the human body, the immune system responds by producing antibodies to proteins and other immunogenic (capable of provoking an immune response) compounds of the pathogen. A number of antibodies are also produced against each immunogenic molecule, each binding to different regions of the molecule. Antibodies, as part of the normal immune response (Figure 6.4), are essential in the protection of the human body from invasion by pathogens. Antibodies increase the ability of immune cells to engulf and destroy bacteria, stimulate the production of chemicals that kill fungi, and neutralize viruses by coating viral proteins that are required for entry into host cells. Furthermore, antibodies neutralize some toxins; this is why people who have deep skin wounds are given an injection of antibodies that specifically bind to tetanus toxin. This binding prevents the toxic action of any toxins produced by Clostridium tetani, a microbe that is able to invade and colonize deep wounds. The specific binding properties of antibodies are invaluable to biotechnologists, particularly as a key part of diagnostic systems. For example, antibodies against flagellar proteins of Salmonella are used in the food industry and in hospitals to definitively identify Salmonella. The process is similar to that described in

Diagnostic Systems

165

Variable regions

Constant region

FIGURE 6.4 Structure of an antibody of the IgG class. There are two light-chain polypeptides and two heavy-chain polypeptides (bolded). The protein is stabilized by disulfide links between polypeptides. The variable regions are responsible for antigen binding and give the antibody its specificity of binding. Each variable region can bind to one antigen molecule.

Figure 6.1, except that the presence of Salmonella is confirmed by an antibody-based test instead of a biochemical test. Latex agglutination tests are popular (Figure 6.5). Antibodies that bind specifically to flagellar proteins of Salmonella are first adsorbed onto latex beads. A drop of solution containing these beads is then added to a drop from a culture that has been tentatively identified as Salmonella. If Salmonella is present, antibodies on the latex beads will bind to Salmonella cells. This will create a network that results in the agglutination of Salmonella. This agglutination is readily visible as a suspension of particles in the drop. Antibody-based systems have been developed for a number of important foodborne pathogens such as E. coli O157:H7, the strain of E. coli that causes hemorrhagic symptoms following infection. Many of these systems do not rely on agglutination for detection; other strategies, such as enzyme-linked immunosorbent assays (ELISAs), are often more sensitive (i.e., they can detect smaller numbers of pathogens or smaller amounts of toxins) and are quantitative (i.e., they allow measurement of the numbers of pathogens or the concentration of toxin). A number of different strategies are possible with ELISA systems. In sandwich ELISAs (Figure 6.6), the antibody is coated on the bottom of wells in plastic multiwell plates (96-well plates are often used). This is the capture antibody. For example,

166

Introduction to Food Biotechnology

Latex bead

Antibody

+

Salmonella

FIGURE 6.5 Latex agglutination assay. Antibodies specific to Salmonella are bound to latex beads. These antibodies link Salmonella cells to a network of latex beads, causing agglutination (clumping).

antibodies that bind specifically to aflatoxin (a potent mycotoxin) could be coated onto a 96-well plate. Samples of cereal grains (after grinding and other pretreatments) could then be added to the wells of the plate. The capture antibody would essentially capture any aflatoxin present in the sample. The plates would then be thoroughly rinsed so that the only substance remaining from the grain sample would be aflatoxin caught by the capture antibody. A second antibody would then be added (the detection or reporter antibody) that binds to a different epitope of the antigen than the capture antibody. This antibody would be covalently bound to a compound that allows detection. For example, it could be linked to an enzyme such as horseradish peroxidase or other enzyme that can convert a substrate into a visible end product. Not surprisingly, then, the next step in the ELISA is to wash off any detection antibody that has not bound to aflatoxin. A chromogenic substrate is then added, to reveal detection antibody–aflatoxin–capture antibody complexes. The amount of colored end product is directly proportional to the amount of aflatoxin present in the original grain sample.

Diagnostic Systems

167

Coat capture antibody onto well Add sample containing toxin

+

Capture antibody binds to toxin

Add detection antibody

Add substrate

Colored end product appears

FIGURE 6.6 Sandwich ELISA for the detection of a toxin. Increased numbers of toxin molecules will lead to increased formation of colored end product.

Another popular strategy is competitive ELISA (Figure 6.7). As an example, consider a competitive ELISA aimed at detecting Salmonella enteritidis. Instead of immobilizing a capture antibody to the wells of an ELISA plate, the target antigen is attached to the plate. The food sample, after enrichment, is then added to the wells, and antibodies against the target antigen are added at the same time. The immobilized antigen and any S. enteritidis cells in the food sample will compete for binding sites on the antibodies. If S. enteritidis is present, fewer antibodies will bind to the immobilized antigen. After the wells are washed, a second antibody is added; this antibody binds to any antibody that is in a particular class. If the first antibody was mouse IgG, the second antibody would bind to the constant region of any mouse IgG molecule. The second antibody is also covalently bound to an enzyme that is able to catalyze a reaction leading to a colored end product. This allows visualization of the amount of immobilized antigen that bound to the first antibody. Thus, low levels of color occur when S. enteritidis is present in the original food sample. Competitive and sandwich ELISA systems are available for the detection of a range of food-borne pathogens; they typically are able to detect 103 to 105 cells per milliliter. Thus, if a 10-g food sample containing less than 104 cells of a pathogen is suspended into 100 mL of buffer, and then tested by ELISA (or an agglutination test), the pathogen will not be detected. Clearly, this is not sensitive enough for most

168

Introduction to Food Biotechnology

Coat plate with antigen from Salmonella

Add sample containing Salmonella, and add antibody

+

Salmonella in sample bind to most of the antibody

Add detection antibody Add substrate

Very little end product appears

FIGURE 6.7 Competitive ELISA for the detection of Salmonella. As the density of Salmonella increases, less antibody is available to bind to the antigen coating the wells. This leads to less binding of detection antibody to antibodies in the wells and fewer colored end products.

food applications. Consequently, enrichment is usually required before the use of ELISA or other immunological assays. 2. Manipulating the Immune Response As we have seen, antibodies are part of the mammalian immune response to the presence of foreign antigens. How can we manipulate this response in order to obtain antibodies that can be used in a diagnostic test? Basically, there are three approaches to the production of antibodies: polyclonal, monoclonal, and recombinant antibodies. Before 1975, the only practical method was to inject microbes or purified components of microbes into animals such as rabbits or mice. After a prolonged period (at least 3 weeks), antibodies could be purified from blood of the animals. This strategy has several disadvantages. Many animals are required for large-scale antibody production, which is expensive. Even with purified compounds (e.g., a specific protein antigen found on the surface of a microbe), this approach yields an antiserum composed of a mixture of antibodies that bind to different epitopes of the antigen. Hence, these are called polyclonal antibodies. Some of these antibodies may have high affinity and specificity, but others may have less ideal binding properties. Antibodies with a low degree of specificity in a diagnostic assay would give a large proportion of false positives, whereas antibodies with low affinity would

Diagnostic Systems

169 Inject mouse with antigen Mouse produces antibodies against antigen Collect spleen cells Mix with myeloma cells PEG

Hybridomas form via fusion of spleen and myeloma cells

Unfused spleen cells die

Plate on HAT medium

Hybridoma cells survive

Unfused myeloma cells die

Screen hybridoma cells for desired clone

FIGURE 6.8 Procedure used to isolate hybridoma cells producing monoclonal antibodies. (PEG = Polyethylene glycol; HAT = hypoxanthine aminopterin thymidine).

be less sensitive (i.e., unable to detect low numbers of pathogens). Finally, antisera are very susceptible to variation from animal to animal. This causes inconsistency in the ability of the antibodies in the antiserum to detect their target. Despite these problems, polyclonal antibodies are sometimes used to develop diagnostic assays. Fortunately, monoclonal antibodies can also be obtained against most antigens. Unlike polyclonal antibodies, which are derived from several lineages (clones) of B-cells, monoclonal antibodies, as their name suggests, arise from only one clone. This means that the antibodies produced by each cell of the clone are identical and bind to the same epitope of the antigen. Thus, there is little batch-to-batch variation in the binding efficiency of monoclonal antibodies. Another advantage is that, once established, monoclonal antibodies are produced by cells grown in culture. Antibodies can be harvested from spent (i.e., used) media of cell cultures and then purified by a number of methods (e.g., column chromatography). Therefore, the production of large amounts of monoclonal antibodies does not require the use of large numbers of animals nor the long incubation times needed to produce polyclonal antibodies. However, animals must be used to isolate clones of B-cells that produce the desired antibody (Figure 6.8). Usually, a mouse is injected with a compound or mixture of compounds derived from the target organism. If the target is a pathogen, one might use purified flagellar protein or another protein that is present on the

170

Introduction to Food Biotechnology

surface of the organism. If the aim is to develop an assay for toxin detection, the toxin itself can be injected, as long as the toxin does not adversely affect the mice. If mice are vulnerable to the toxin, the toxin can be inactivated by chemical treatment (e.g., denaturation with formaldehyde). With luck, this will result in loss of the biological activity of the toxin but maintenance of its three-dimensional structure. The binding of antibody to antigen is based on structural interactions, and conservation of this structure is essential. Several weeks later, the mouse is killed and its spleen is removed. Spleen cells are then mixed with myeloma cells — tumor cells that are able to grow indefinitely in cell cultures. Polyethylene glycol (PEG) is then added to the mixture. Recall from Chapter 4, Section II.D.2, that PEG induces the fusion of plant protoplasts. It has a similar effect on animal cells, and cell fusion occurs in the mixture of myeloma and spleen cells. The aim at this stage is to fuse B-cells from the mouse spleen with myeloma cells. The resulting cells (hybridomas) will be able to produce antibodies (derived from the B-cells) and grow indefinitely in culture (derived from the myeloma cell). Unfused B-cells cannot be used directly to produce monoclonal antibodies, because they are not immortal; they die after several weeks in culture. Undesirable fusion products also occur. Myeloma cells fuse with myeloma cells and B-cells fuse with B-cells. Desired hybridoma cells are isolated using an elegant selection scheme. The myeloma cells used are mutants lacking a crucial enzyme that normally allows cells to import hypoxanthine, a compound that can be converted into nucleotides required for DNA synthesis. However, this enzyme is present in Bcells. After fusion, the cells are grown in a medium that forces them to depend on hypoxanthine in the medium as a source of nucleotides. The only cells that survive and grow will be B-cells and hybridomas. B-cells that do not fuse with myeloma cells do not survive for very long, because of their limited ability to divide. The surviving hybridomas are diluted and seeded, so that individual cells are placed in a separate well (e.g., one cell per well of a 96-well plate). Each hybridoma cell then divides, giving a suspension of identical progeny cells. Thus, a series of clones is produced, each of which is able to divide indefinitely in culture and produce only one kind of antibody. This is a crucial point: each B-cell produces only one type of antibody. Therefore, if a clone is transferred into a larger culture vessel, the cells will continue to divide and produce monoclonal antibody indefinitely. These antibodies have a single specificity. This means that all antibodies produced by a particular clone recognize and bind to the same epitope. At this point, it is necessary to screen the clones to identify the clones that produce useful antibodies. One screening method involves ELISA; the antigen is immobilized onto a multi-well plate (this is often done through simple absorption), and spent medium from each clone is added to each well. Each sample of spent medium contains antibodies. After rinsing, useful clones can be identified through the addition of an enzyme-linked secondary antibody (e.g., if mice were used for the primary immunization, we could use an anti-mouse antibody produced by a goat). The valuable clones are transferred to larger culture vessels, and large amounts of antibodies can then be collected from the culture medium. The immortal properties of the myeloma cell allow these antibody-producing cells to be produced indefinitely.

Diagnostic Systems

171

One problem associated with the production of monoclonal antibodies is that mammalian cells are much more difficult to grow than most bacteria and fungi. Nevertheless, many biotechnology companies have successfully marketed diagnostic systems based on monoclonal antibodies. The third option for obtaining antibodies is through gene cloning methods. Recombinant antibodies can be obtained by isolating mRNA from immunized or nonimmunized mouse cells. If the mouse is not immunized, B-cells with a broad range of specificities are obtained, whereas an immunized mouse will provide a more restricted range of antibody specificities. Once the mRNA is isolated, cDNA is made using reverse transcriptase, and PCR is used to amplify antibody genes. The next step is usually to insert these genes into a phage vector, creating a library of antibody genes. If the vector is constructed so that recombinant proteins are produced from the antibody genes, then the library can be screened for the desired antibodies. Once the phage containing the desired antibody is isolated, it can be used to produce large amounts of monospecific antibodies, much like monoclonal antibodies. There are several advantages to the recombinant approach. It is fast and extremely flexible. Antibody genes can be directly manipulated to find, for example, the smallest amino acid sequence that confers specific recognition of the antigen. Also, because the antibody genes are already cloned, sequencing is straightforward; this may allow design of new sequences that have improved characteristics (e.g., improved specificity or greater heat stability). 3. Future Applications of Immuno-Assays There is currently great interest in the development of immuno-assays, as seen in Table 6.3, which was compiled from a search of the Current Contents™ database for the period January 2000 to July 2001, using the keywords “food” and “immunoassay”). To make this survey representative of current diagnostic development goals, we excluded studies that simply used currently available immuno-assays to detect residues in food and studies that were veterinary in scope (e.g., detection of brucellosis in cattle). Of the 76 studies, 33 were aimed at the detection of specific pesticides in food, soil, or water samples. Eight studies attempted to develop immuno-assays for the detection of mycotoxins and six for the detection of antibiotic residues. Surprisingly, only nine studies targeted detection of bacterial pathogens or toxins. Five studies were aimed at assessment of food quality (not related to pathogens). To summarize, the current thrust of diagnostic development appears to be related to the detection of undesirable residues in food; another trend is toward the development of diagnostic techniques that can be incorporated into an electronic detection system. These are biosensors and will be further discussed in the next section. Detection of pesticides, antibiotics, and mycotoxins using antibodies is difficult because these compounds are usually low-molecular-weight nonprotein compounds. These sorts of compounds tend to have low immunogenicity (they do not elicit antibody production). However, if they are conjugated to other, larger compounds, their immunogenicity often increases dramatically. The small compound is said to act as a hapten in this context.

172

Introduction to Food Biotechnology

TABLE 6.3 Diagostic Immuno-Assays under Development Detection Goal Pesticide Mycotoxins

Number of Studies 33 8

Antibiotic residues Salmonella spp.

6 6

Allergen

5

Meat identification Staphylococcal enterotoxin Vitamins Plant toxins Chitin oligos? Heat treatment of milk Lipid oxidation in meat Listeria spp. Spoilage microbes Gluten Transgenic crops Algal toxins Norwalk virus

3 2 2 2 1 1 1 1 1 1 1 1 1

Types of Immuno-Assay Agg,a cELISA,b biosensor, ELISA, fluorescence, biosensor, cELISA, immuno-affinity Biosensor, cELISA cELISA, ELISA, automated assay, biosensor, agg ELISA, luminescence, biosensor, dot blots cELISA, ELISA Biosensor, fluorescence Biosensor cELISA, fluorescence cELISA cELISA cELISA ELISA ELISA ELISA ELISA Biosensor ELISA

a

agg, agglutination. cELISA, competitive ELISA. Note: Current Contents™ was searched using the keywords “food” and “immuno-assay” for the period January 2000 to July 2001. b

Why bother to develop immuno-assays for these compounds, given the problems developing antibodies specific for them? Conventional methods of pesticide detection rely on time-consuming methods such as HPLC or gas chromatography (GC), which require expensive instrumentation. This limits the number of laboratories that are equipped to detect pesticide residues and the number of samples that can be tested. Consequently, it is difficult to resolve the current high level of consumer uncertainty about the safety of pesticide residues in food. Also, government regulators throughout the world are concerned with the sporadic occurrence of high levels of pesticide residues, particularly on fruits or vegetables that are consumed without processing. The development of easy, rapid diagnostic tests for pesticides would allow more frequent testing of food commodities and more efficient tracing of pesticide residues as they enter the soil and water environment.

Diagnostic Systems

F. LUMINESCENCE

173 AND

DIAGNOSTICS

1. Hygiene Assessment Many food processors would like to be able to monitor the level of microbial contamination of surfaces and equipment. For example, the adequacy of equipment cleaning practices can be assessed and monitored by measuring the microbial load of the equipment after cleaning. This can be done by traditional cultural methods, but this approach suffers from the same problems described earlier in this chapter. Namely, it is laborious, time consuming, and expensive. Recently, several biotechnology companies have marketed devices that allow the measurement of adenosine triphosphate (ATP), an indicator of microbial load. This approach was pioneered by microbial ecologists, who have long grappled with the problems associated with measurement of microbial biomass and activity in natural systems such as soils and sediments. The ATP approach to measuring microbial biomass is possible because all organisms use ATP as the energy currency of the cell. To grow, microbes must take in energy sources (e.g., carbohydrates such as glucose) from their environment and convert this energy into ATP. The resultant ATP is then used to fuel a multitude of processes, such as cell wall growth, protein synthesis, and membrane formation. Because all cells use ATP, the presence of ATP on an inert surface or equipment component indicates the presence of microbes. The detection of ATP tells the user nothing about the identity of the organism (it could be Salmonella, yeasts, food particles, human saliva, etc.); it is considered to be only an indicator of hygiene. Significant levels of ATP on a surface, for example, may indicate that the cleaning process is inadequate and must be revised. This is particularly useful in dairy operations, where hygiene is critical and visual monitoring of hygiene is insufficient. The detection of ATP typically involves taking a swab of the area of concern and the lysis of any swabbed microbial cells. If ATP is present, an enzyme called luciferase (isolated from fireflies), causes the emission of light, through the following reaction: luciferin + ATP + Mg 2+ luciferase  → oxyluciferin + AMP + CO 2 + light The amount of light produced is directly proportional to the amount of ATP. Because ATP detection relies on the sensing of low levels of light, specialized instruments (luminometers) are required to monitor hygiene using ATP. 2. Novel Applications of Luminescence Luminescence has great potential in other areas of food safety assurance. Most of these applications involve engineered use of bacterial luciferase genes, which have been successfully transferred to Bacillus spp., Listeria spp., Staphylococcus spp., Aeromonas spp., and lactic acid bacteria (e.g., Lactococcus lactis). Although the lightemitting reaction catalyzed by bacterial luciferase is different from the eukaryotic

174

Introduction to Food Biotechnology

reaction (one key difference is that ATP is not used directly), it is similar in that the intensity of light emission is related to the cellular viability and energy status. This has led to novel applications of recombinant strains of bacteria relevant to food. Bacteria with recombinant luciferase genes can be used to assess the efficacity of cleaning regimes in industrial food processing operations. Luminescent bacteria are applied to preparation surfaces, cleaning is completed, and bacterial light emission (if present) is detected. This allows an assessment of efficacity against a specific microbe in real time (virtually immediate); in contrast, a conventional plate count assessment would require at least 24 h. Another example of luminescence technology is the use of recombinant phage with the bacterial luciferase gene. This is an exciting technology because each bacterial species has a set of specific phage that are unable to infect other bacteria. Some phage have a broader host range, which can be useful. For instance, a phage that infects bacteria within the Enterobacteriaceae family would be a useful monitor of a range of food-borne pathogens, including E. coli and Salmonella. Such a phage could be incubated briefly (1 h) with a food sample, and if enteric bacteria were present, they would be infected by the phage, resulting in expression of phage genes, including luciferase. This would result in detectable light emission.

G. BIOSENSORS 1. Applications of Biosensors The term “biosensors” has been used to describe a number of distinct diagnostic systems; this has made it difficult to define the term precisely. However, we can make the following generalization: biosensors have a biological sensor that is connected to a transducer. A transducer is a device capable of converting signals from the biological sensor into signals (usually electrical) that can be easily recorded and stored. For example, a number of biosensors use the specificity of antibody–antigen binding to detect pathogens in food samples. When the pathogen is present, it binds to the antibody. The key event follows: binding of antigen to antibody produces an electrical signal that can be detected and recorded. Biosensors, then, are an example of what science fiction authors describe as “cybernetics” — the fusion of organic matter to electronic circuitry. Although less dramatic than such fictional constructs as the “Borg” of Star Trek fame, biosensors offer many examples of inspired cooperation among microbiologists, biochemists, physicists, and electronic engineers. Biosensors have many applications in clinical settings (e.g., diagnosis of foodborne pathogens from stool and other samples) and in maintenance of food safety (e.g., assessment of microbial loads or detection of specific pathogens in food). For example, it is theoretically possible to design biosensors that are sensitive (e.g., able to detect one pathogen in 25 g of food), selective (able to discriminate specific pathogens from a large background of nonpathogenic microbes), fast (real time), automated, portable, and inexpensive. To date, this potential has not been realized, but research in biosensor development is highly active and steady improvements in design are predicted.

Diagnostic Systems

175

Biosensors also have many applications that are not related to detection of pathogens. For example, one of the problems with large-scale cultivation of microbes (see Chapters 7 and 8) is that it is difficult to monitor certain aspects of microbial growth. The rate of formation of a desired product often can be monitored only through the analysis of samples taken from the bioreactor (vessel used to grow large microbial cultures). It would be highly preferable to monitor product formation continuously, perhaps by having a product-specific electrode in the interior of the bioreactor. Biosensors are available for certain products (e.g., lactic acid) and a number of other bioreactor parameters (e.g., glucose consumption, cell growth, and viability). Continuous monitoring is also useful for many food safety or spoilage applications, particularly in the processing of liquids (milk, beer, etc.), where it is desirable to monitor microbial numbers in line (in piping systems used to transfer liquids from one vessel to another or to packaging processes). For example, postpasteurization contamination of beverages and foods is a significant cause of spoilage and has been implicated in outbreaks of food-borne illness. In 1987, postpasteurization contamination by S. enteritidis of milk used to make ice cream led to one of the largest recorded outbreaks of food-borne illness in the U.S. In-line detection could prevent this sort of accident, as well as the more common problem of spoilage arising from post pasteurization contamination (e.g., reduction of shelf life of milk by postpasteurization contamination by psychotrophic microbes). Contamination of processing equipment is often difficult to eradicate. A continuous monitoring system that detects microbial growth in the lines is much better than monitoring based on examination of discrete samples or equipment swabs. If microbial growth could be immediately detected in transfer lines, the process could be stopped and the contamination eliminated, thus avoiding the production of large amounts of contaminated product. Conventional diagnostic systems based on enrichment and selective culture are not adaptable for continuous monitoring. Instead, they require the collection of discrete sampling units (batch samples). Each sample is then cultured in the appropriate media. Continuous monitoring using a culture-based system would require an infinite (or at least very large) number of sampling units, whereas biosensors can continuously monitor without the collection of discrete samples. Biosensors can also be used for batch sampling. One important application of biosensors is to speed up pathogen identification using culture-based methods. For example, spoilage of fresh meat is an economically important problem and is often linked to the presence of high levels of a variety of spoilage microbes. Conventional monitoring is done through plate counting (total bacterial count) after incubation on nonselective media that allow a wide range of bacteria to grow. However, this requires at least 24 h — not ideal for preventing meat spoilage. Biosensors have been designed that assess levels of microbial contamination after short (1 h or less) incubation of meat samples. It is also often desirable to continuously monitor physical processes in the food industry. Physical parameters such as temperature can easily be monitored continuously, allowing immediate adjustment if the temperature strays outside a defined range, and also providing a record of temperature changes. A continuous record can

176

Introduction to Food Biotechnology

be useful if product quality declines; deviations in the temperature of the process could be a causal factor. Biosensors can be used to monitor some physical–chemical processes (e.g., CO2 production). 2. Types of Biosensors Biosensors can be classified according to the type of sensor, the transduction strategy, and the directness of the assay. Affinity-based biosensors rely on specific recognition between the sensor and the target. Antibodies are most commonly used, but nucleic acid hybridization, similar to that used in gene probe assays, and receptor–ligand interactions (e.g., insulin binding to insulin receptor molecules) are also used to create affinity-based biosensors. Antibody-based biosensors have been developed for most of the major food-borne pathogens (e.g., Salmonella, E. coli O157:H7, and L. monocytogenes). How is antibody–antigen binding detected and converted into an electrical signal? One approach is to immobilize antibodies to the surface of piezoelectric crystals, which are very sensitive to changes in mass. When antigens bind to the antibody, they increase the mass of the antibody–crystal complex. Piezoelectric crystals are unusual in that the application of external forces (e.g., gravity pulling on an attached antibody) leads to oscillation and a detectable electrical potential. When an antigen binds to the antibody, it increases the mass of the complex, which changes the frequency of oscillation of the crystal. This change of frequency can be detected electronically. Enzyme-linked antibodies can also be used in biosensors. For example, a biosensor that detects S. aureus uses antibodies immobilized to an electrode (Figure 6.9). These antibodies “catch” the bacteria; antibodies that bind to another epitope of S. aureus are then added. These antibodies are covalently linked to horseradish peroxidase (HRP). The electrode is then moved to a solution containing amino salicylic acid (AMSA). Another enzyme (glucose oxidase) is also immobilized onto the electrode; the sole purpose of glucose oxidase is to generate hydrogen peroxide from glucose and oxygen. HRP catalyzes the reaction between hydrogen peroxide and AMSA to form 5-ASA-quinoneimine (ASAQ), which is then reduced by an electron supplied by the electrode. This current of electrons to ASAQ is detectable electronically. One of the chief problems with affinity biosensors has been regenerating electrodes between samples. All bound antigens must be removed from the electrode to restore its sensitivity. It is difficult to achieve this without the use of harsh chemicals (e.g., 8 M urea) that tend to decrease the longevity of the electrode. This is one of the reasons for increased interest in affinity biosensors based on nucleic acid hybridization. Unlike antibody–antigen binding, hybridization occurs over a relatively long molecular distance and is stabilized by frequent hydrogen bonds. In contrast, antibody–antigen binding typically occurs within short amino acid sequences and involves a mixture of hydrogen bonds, ionic attractions, and nonpolar interactions. Besides stability, nucleic acid hybridization is also attractive because it can easily be undone through adjustment of the ionic environment of the solution. The design of affinity biosensors using nucleic acid hybridization usually

Diagnostic Systems

177

A

Electrode

*

B

C

* *

*

Another view of the same electrode D O2 Glucose oxidase

H2O2+ AMSA

*

*

ASAQ e

-

FIGURE 6.9 A biosensor for the detection of S. aureus. Cells of S. aureus (A) bind to antibodies on the electrode (B). A second antibody, which binds to a different epitope on the cell, is added (C). This antibody is covalently linked to horseradish peroxidase (HRP), designated as “*”. The electrode is then removed and placed in a solution of amino salicylic acid (AMSA). (D). The electrode is also coated with glucose oxidase. This enzyme generates hydrogen peroxide (H2O2). HRP catalyzes the reaction between AMSA and H2O2 to form 5-ASA-quinoneimine (ASAQ). An electron from the electrode then reduces ASAQ. Current flow from the electrode increases as the numbers of S. aureus increase.

involves immobilization of oligonucleotide probes onto an electrode. Hybridization can be detected by similar interfaces as with antibodies (e.g., piezoelectric crystals that react to the change of mass induced by hybridization with target nucleic acids).

178

Introduction to Food Biotechnology

Optical sensors have also been successfully used. These sensors rely on changes in the optical qualities of substances coating an electrode. For example, hybridization might cause a change in the refractive index of a coating substance, which is then detected by the use of sensitive light detectors. Another group of biosensors relies on microbial metabolism to monitor microbial growth or levels of microbial contamination. The most common strategy is to use oxidoreductases in cells as a signal of microbial growth. Usually, oxidoreductases change the structure of a mediator chemical that in turn triggers a response from the transducer. In one system, bacteria are immobilized onto an electrode that is coated with p-benzoquinone, a mediator. Dehydrogenase, a type of oxidoreductase found in most microbial cells, transfers an electron to the mediator. The reduced mediator then passes on this electron to the electrode, resulting in a detectable current. Although effective, this and other biosensors based on microbial metabolism function well only when large densities of microbes are present. This may limit their use in food sampling to circumstances in which low levels of contamination are acceptable but high levels are not (e.g., fresh meat). Biotechnologists have also been successful in the design of biosensors that detect changes in metabolites such as glucose or lactate. Changes in the concentration of glucose can be continuously monitored by biosensors that use the enzyme glucose oxidase. This enzyme catalyzes the following reaction: oxidase glucose + oxygen glucose  → H 2O2 + gluconate

The key part of a glucose biosensor that uses glucose oxidase is a system for detecting changes in oxygen or H2O2 levels. Several strategies are possible; many use fluorescent compounds to detect changes in oxygen. One design uses tris (1,10 phenanthroline) ruthenium chloride, a compound that has different fluorescence characteristics when exposed to oxygen. This biosensor has a fiber optics system that allows excitation of the ruthenium and detection of the resulting fluorescence. Changes in glucose are important in the production of glucose syrups from starch. The conversion of starch to glucose is a complex process driven by a number of microbe-derived enzymes, and the concentration of glucose in starch undergoing processing is an important parameter. Microbial production of food-related metabolites is another process that can be monitored by the use of glucose biosensors, because many of the feeds used in bioreactors are rich in glucose. Glucose is also an important food ingredient in many processed foods (e.g., candies and other confections). The ability to monitor glucose without setting up extensive analytical facilities is attractive as a means to efficiently maintain quality control. Despite the utility of glucose biosensors in food biotechnology, the driving force behind their development has been the enormous market of diabetics who need to frequently monitor their blood glucose. Recent clinical studies have demonstrated that tight control over blood glucose, achieved through frequent testing, results in a lower incidence of health complications. This has increased the need for easy, quick monitoring systems, and has also driven research into implantable systems that ultimately will lead to an artificial pancreas. This would consist of a subcutaneous

Diagnostic Systems

179

biosensor (probably using glucose oxidase) that continuously monitors blood glucose, connected to an electronic system that controls operation of an insulin pump. Such a device would create a “closed loop,” so that insulin-dependent (type I) diabetics would not need to monitor blood glucose through skin pricks nor inject insulin. Achieving this goal will require significant improvements in biosensor design, which will probably be adaptable for use in the food industry. Improvements in our ability to miniaturize electronic components and to develop “laboratories in a chip” are also expected to lead to more frequent application of biosensors in the food industry. Applications are currently not common, mostly because of low sensitivity, interference by compounds in food matrices, and difficulties regenerating electrodes. The potential benefits, though, are significant enough to justify continued research and development of these elegant diagnostic systems.

RECOMMENDED READING 1. de Boer, E. and Beumer, R. R., Methodology for detection and typing of foodborne microorganisms, Int. J. Food Microbiol., 50, 119, 1999. 2. Ray, B., Fundamental Food Microbiology, CRC Press, Boca Raton, FL, 1996. 3. Karch, H., Bielaszewska, M., Bitzan, M., and Schmidt, H., Epidemiology and diagnosis of shiga toxin-producing Escherichia coli infections, Diagn. Microbiol. Infect. Dis., 34, 229, 1999. 4. Curry, A. and Smith, H.V., Emerging pathogens: Isospora, Cyclospora and microsporidia, Parasitology, 117, S143, 1998. 5. Reiff, F. M., Roses, M., Venczel, L., Quick, R., and Witt, V. M., Low-cost safe water for the world: a practical interim solution, J. Public Health Policy, 17, 389, 1996. 6. Herwaldt, D. L., Ackers, M.-L., and the Cyclospora Working Group, An outbreak in 1996 of cyclosporiasis associated with imported raspberries, New Engl. J. Med., 336, 1548, 1997. 7. Mortimore, S. and Wallace, C., HACCP: A Practical Approach, Chapman & Hall, London, 1994. 8. Savage, R. A., Hazard analysis critical control point — a review, Food Rev. Int., 4, 575, 1995. 9. Mozola, M. A., Detection of microorganisms in foods using DNA probes targeted to ribosomal RNA sequences, Food Biotechnol., 14, 173, 2000. 10. Lantz, P.-G., Hahn-Hägerdal, B., and Rådström, P., Sample preparation methods in PCR-based detection of food pathogens, Trends Food Sci. Technol., 5, 384, 1994. 11. Bellin, T., Pulz, M., Matussek, A., Hempen, H. G., and Gunzer, F., Rapid detection of enterohemorrhagic Escherichia coli by real-time PCR with fluorescent hybridization probes, J. Clin. Microbiol., 39, 370, 2001. 12. Talary, M. S., Burt, J. P. H., and Pethig, R., Future trends in diagnosis using laboratoryon-a-chip technologies, Parasitology, 117, S191, 1998. 13. Umek, R. M., Lin, S. W., Vielmetter, J., Terbrueggen, R. H., Irvine, B., Yu, C. J., and Kayyem, J. F., Yowanto, H., Blackburn, G. F., Farkas, D. H., and Chen, Y. P., Electronic detection of nucleic acids: a versatile platform for molecular diagnostics, J. Mol. Diagn., 3, 74, 2001. 14. Lee, H. A. and Morgan, M. R. A., Food immuno-assays: applications of polyclonal, monoclonal and recombinant antibodies, Trends Food Sci. Technol., 4, 129, 1993.

180

Introduction to Food Biotechnology

15. Paraf, A., A role for monoclonal antibodies in the analysis of food proteins, Trends Food Sci., 3, 263, 1992. 16. Spinks, C. A., Broad-specificity immuno-assay of low molecular weight food contaminants: new paths to Utopia! Trends Food Sci. Technol., 11, 210, 2000. 17. Walker, A. J., Jassim, S. A. A., Holah, J. T., Denyer, S. P., and Stewart, G. S. A. B., Bioluminescent Listeria monocytogenes provide a rapid assay for measuring biocide efficacy, FEMS Microbiol. Lett., 91, 251, 1992. 18. Kodikara, C. P., Crew, H. H., and Stewart, G. S. A. B., Near on-line detection of enteric bacteria using lux recombinant bacteriophage, FEMS Microbiol. Lett., 83, 261, 1991. 19. Deshpande, S. S., Principles and applications of luminescence spectroscopy, Crit. Rev. Food Sci. Nutr., 41, 155, 2001. 20. Ivnitski, D., Abdel-Hamid, I., Atanasov, P., and Wilkins, E., Biosensors for detection of pathogenic bacteria, Biosensors Bioelectron., 14, 599, 1999. 21. Rogers, K. R., Principles of affinity-based biosensors, Mol. Biotechnol., 14, 109, 2000. 22. Jaremko, J. and Rorstad, O., Advances toward the implantable artificial pancreas for treatment of diabetes, Diabetes Care, 21, 444, 1998.

7

Cell Culture and Food I. OVERVIEW

Cell culture has facilitated basic research into animal, plant, and microbial cell metabolism. The ability to grow uniform cultures of cells is particularly useful for the study of cellular metabolism and behavior outside of the complex multicellular environment of a plant or animal. This is relevant to food; for example, the study of human intestinal epithelial cells has elucidated the mechanisms of nutrient digestion and uptake. Such studies would be difficult or impossible in intact animals. With regard to unicellular organisms, cell culture has been equally important, and the study of isolated species or strains of bacteria and fungi has provided great insight into basic microbial structure and function. This is also relevant to food. For example, discovering the nature of toxins produced by Escherichia coli O157:H7 has helped us understand the characteristics that make this strain so much more dangerous than other strains of E. coli and has given us an important diagnostic tool — that is, instead of trying to detect the bacterium, in some cases it may be easier to detect the toxin. A wide range of cell culture applications are directly relevant to food production and processing. Many of these, such as the use of yeasts to ferment foods into ethanol, have been employed by humans throughout recorded history and are often called traditional biotechnology. Others, such as the use of bacteria to produce flavor enhancers such as glutamate, have existed only in the 20th century. In this chapter, we examine both traditional and modern microbial food biotechnologies. The brewing and dairy industries, the main bastions of traditional biotechnology, have widely embraced many aspects of what is usually considered modern biotechnology. For example, both industries use advanced control and diagnostic systems in bioreactor design and both have made use of recombinant DNA technology to modify cells. Recombinant yeast and bacteria are technically ready to be used by the dairy and brewing industries, but are currently awaiting public acceptance before widespread introduction. Consequently, it is important to include these traditional industries in a discussion of modern biotechnology. Furthermore, they are among the most successful biotechnologies. Economically, the brewing industry and associated distilling and wine-making industries constitute a greater force than all other biotechnologies. For example, in Canada, biotechnology companies (excluding brewing and dairy companies) had a total revenue of $1.67 billion in 1993. In 1996, Canadians spent $8.8 billion on beer alone. In the U.S., 2.5 million people are employed in the brewing industry, and in Germany, 108 million hectoliters of beer were consumed in 1996. These figures demonstrate that traditional yeast fermentation is an important economic force globally. Other biotechnologies, especially those

181

182

Introduction to Food Biotechnology

applied to human therapeutics, are likely to increase in value, closing the economic gap between them and traditional biotechnologies. For example, in 1997, global revenues from recombinant therapeutic proteins were $17 billion. Nonetheless, traditional biotechnologies will remain a potent economic force in the future. Yeast cells ferment carbohydrates to ethanol, forming the basis of the brewing industry; lactic acid bacteria (LAB) ferment carbohydrates to lactic acid, a necessary part of the production of cheese, yogurt, and various other dairy products. Both ethanol and lactic acid are end products of energy-generating fermentation pathways. However, other aspects of microbial metabolism have been exploited in the food industry. Microbial enzymes are used in a number of food processes, including starch processing, where amylases convert starch into glucose and other useful compounds, and cheese making, where lipases are often used to improve cheese flavor. Microbial polysaccharides (e.g., xanthan gum) are widely used in the food industry as thickeners and texture modifiers. Amino acids (e.g., glutamate) derived from microbial cells are also important. Finally, microbial biomass is sometimes used directly as food. Yeast biomass is an economically important product, used for bread rising and also directly as a food (e.g., Vegemite™). Meat substitutes can also be made from microbial biomass — Quorn™, made from biomass of Fusarium graminearum, is an example, as is tempeh, which is composed of soybeans and biomass of Rhizopus oligosporus. The primary objective in this chapter is to explain how cell culture can be used to produce these valuable products. Another aim is to develop an understanding of the strategies that can be used, and that will be used in the future, to improve the efficiency of production of cellular products. Although animal cell products (e.g., antibodies) have application to the food industry, and plant cell products (e.g., flavor compounds such as vanillin) may be important in the future, we will focus on microbial cell products, because these have so far been the most useful to the food industry (Table 7.1).

II. BREWING A. INTRODUCTION Most cultures have at least one traditional alcoholic beverage that is produced from grains, vegetables, or fruits. Usually Saccharomyces cereviseae is the driving force behind the production of ethanol. Although it is immensely popular throughout the world, ethanol is also very controversial; consumption of alcoholic beverages is forbidden in certain religions, and there is no doubt that abuse of ethanol-containing beverages can lead to addiction and tragic consequences. On the other hand, millions of people enjoy alcoholic beverages without any negative effects, and indeed, evidence is accumulating that the regular consumption of small amounts of alcoholic beverages (e.g., red wine) can be beneficial to human health. Beer is the most widely consumed alcoholic beverage — approximately 1.2 billion hectoliters (1 hL = 100 L) of beer are consumed annually worldwide. Conventional beer brewing makes elegant use of the biological properties of three organisms: S. cereviseae, barley (Hordeum sativum), and hops (Humulus lupulus).

Cell Culture and Food

183

TABLE 7.1 Microbial Cells Used to Make Products that Are Valuable to the Food Industry Product

Use

Producing Organism

Type of Organism

Beer Wine Vodka Baking yeast Yogurt Buttermilk Cheese Glutamate Lysine Glucoamylase Glucose isomerase β-amylase Invertase Xanthan gum Pullulan

Beverage Beverage Beverage Bread making Food Beverage Food Flavor enhancement Feed/food supplement Starch processing Fructose production Starch processing Starch processing Thickener Thickener

Saccharomyces cereviseae S. cereviseae S. cereviseae S. cereviseae Lactobacillus bulgaricus Lactobacillus acidophilus Streptococcus thermophilus Corynebacterium glutamicum Brevibacterium spp. Aspergillus niger Streptomyces spp. Bacillus subtilis S. cereviseae Xanthomonas campestris Aureobasidium pullulans

Fungus (yeast) Fungus (yeast) Fungus (yeast) Fungus (yeast) Bacterium Bacterium Bacterium Bacterium Bacterium Fungus Bacterium (actinomycete) Bacterium Fungus (yeast) Bacterium Fungus

Receive malted barley

Filter mash through spent grains

Grind barley Boil, and add hops Add water and heat Cool, and pitch in yeast

Mash at 65˚C adjuncts

fermentation

Age in tanks

Filter and/or pasteurize

bottle

FIGURE 7.1 Overview of the brewing process.

The overall strategy and sequence of events in the beer-brewing process has not changed in the last century; however, the details of the process and the extent of the brewer’s control over the process have changed enormously. We will examine the brewing process as well as biotechnological approaches to improvements in the brewing process. Brewing can be divided into four discrete stages (Figure 7.1): malting, mashing, fermentation, and packaging. During this process, starch in the barley seed is converted into fermentable sugars (malting and mashing), which are then fermented

184

Introduction to Food Biotechnology

by brewing strains of S. cereviseae into ethanol. Fermentation is usually divided into two stages: the main fermentation, which results in rapid formation of ethanol, and secondary fermentation, wherein the beer matures and improves in flavor. Packaging usually involves filtration and carbonation of the beer, followed by injection of the beer into bottles or cans. Of course, beer is not the only fermented beverage. Wine is also a fermented product, and brandy, whisky, vodka, and rum are all examples of distillates of fermented beverages. The overall process of fermentation is similar in these products. The events prior to fermentation, however, vary widely. In general, beverages that are made from starch-rich substrates require enzymatic pretreatment, because S. cereviseae is incapable of breaking down starch. In contrast, grapes do not require pretreatment because they are rich in sucrose and other carbohydrates that can be directly fermented by S. cereviseae. Before proceeding further, it is worthwhile to consider the range of beer types that consumers enjoy. We can define beer as a fermented beverage made using yeast and malted grains, but this does not come near to describing the incredible range of beverages that are referred to as “beer.” If we restrict our discussion to North American and European beers, we can identify the following major groups: • Lagers are made by bottom-fermenting strains of S. cereviseae. Lager fermentation typically occurs at cool temperatures (4 to 12°C). Lagers are quite different in North America and Europe. North American lagers tend to have little hops, are not bitter, and are made from a mixture of barley, corn, and rice malts. In contrast, European lagers tend to be more bitter, because of higher levels of hops. Lagers usually have a long (up to 3 weeks) period of maturation (secondary fermentation) before bottling. • Ales are made from top-fermenting strains of S. cereviseae. These are fermented at a warmer temperature (12 to 18°C) and typically have a short maturation period (e.g., 1 to 3 d). Because of the warmer fermentation temperature, ales often have a more complex range of flavors than lagers (e.g., fruity flavors arising from yeast-derived esters). The many different styles of ales include strong ales (alcohol content >5%), stouts, porters, pale ales, and bitters. These ales vary in color from yellow to black and have a large range of flavor depth and complexity. • Dry and light beers can be made from lagers or ales, although lagers are more commonly used. They typically have a simpler taste than other beers, because of low levels of nonfermentable carbohydrates in the final product. Both styles are becoming increasingly popular, particularly in North America. The basic difference between dry and light beers is that light beer initially contains fewer fermentable substrates. This results in less alcohol production and fewer calories. Dry beers, in contrast, have similar alcohol and calorie contents to conventional beer but have different sensory qualities. In particular, they are perceived to be less filling than conventional beer.

Cell Culture and Food

185

Seed imbibes water H2O

Aleurone Endosperm

Embryo

Gibberellins are released from the embryo

Gibberellins trigger the aleurone layer to release amylases

Amylases mobilize starch in the endosperm

FIGURE 7.2 Germination of a barley seed. After seed imbibition, gibberellins (circles) diffuse from the embryo to the aleurone layer (outer layer of the endosperm). The aleurone layer responds by releasing amylases (triangles) and other degradative enzymes. This initiates starch breakdown, releasing glucose (hexagons) and other carbohydrates for use by the embryo.

This array of products can be made from the same basic set of ingredients because of: (1) inherent flexibility in the traditional brewing process (e.g., the degree of roasting that occurs during kilning has dramatic effects on beer color and flavor); (2) variation in metabolism of yeast strains; and (3) biotechnological refinements to the brewing process (e.g., use of extra enzymes during mashing).

B. MALTING Brewing begins with malting, and malting begins with the addition of water to barley grains (seeds). This is referred to as steeping. The grains soak in the water and are then spread out on the floor of a humidity-controlled room for at least 48 h. The seeds imbibe water, which triggers germination. Several crucial events occur during germination (Figure 7.2), culminating in the release of degradative enzymes, including α- and β-amylases, proteinases, glucanase, and phosphatase. From the seed’s perspective, the purpose of this release of degradative enzymes is to mobilize carbohydrates, amino acids, and phosphate from the endosperm, the major storage organ of the seed. These nutrients are then used by the growing embryo to fuel the

186

Introduction to Food Biotechnology CH2OH 6

CH2OH O

5 4 OH

3

O 1

OH 2

OH OH

OH R

α-D-glucose

O

O

OH

OH N

CH2OH

OH OH OH

N

R

Maltose

FIGURE 7.3 Structure of α-D-glucose and maltose. Each molecule has a reducing end (R) and a nonreducing end (N). The arrow indicates the position of the α-1,4 linkage in maltose. The numbering convention for carbons in glucose is shown. According to convention, “1” is given to the most oxidized carbon.

initial production of shoot and root tissue. For the brewer, these events are fortuitous, because the carbohydrates in the endosperm will ultimately be one of the major sources of energy and carbon for yeast cells during fermentation. However, carbohydrates are stored in the endosperm in the form of starch, and S. cereviseae, the yeast used in brewing, is unable break starch into fermentable sugars such as glucose or maltose (Figure 7.3). Starch must be enzymatically broken down before S. cereviseae can use it. Barley also contains β-linked carbohydrates such as cellulose and glucans that are less affected by conventional malting or mashing. S. cereviseae cannot use these carbohydrates, so they often persist after fermentation and are present in the finished product. In some cases, these carbohydrates are desirable, because they add body to the beer. However, when the desired beer is light in flavor and body, glucans in particular are troublesome. This is one reason why brewers in North America often replace a proportion of the barley malt with the endosperm of rice or corn, which has low levels of β-glucans. It is clear, then, why the release of amylases from the endosperm is crucial to brewing. Equally important is control over the timing of the enzymatic breakdown of starch. The maltster usually does not let germination proceed past 48 h. The main reason for this is that continued germination results in unacceptable losses of carbohydrates, driven by vigorous growth of the embryo. Kilning the seeds prevents this loss. Dry heat is applied, driving moisture from the seeds. This kills the embryos but will not damage the amylases and other enzymes, as long as the seeds do not become excessively hot (>60°C). Once most of the water has been driven from the seeds, the malted barley is very stable — it can be stored for long periods, providing that the relative humidity of the storage environment is low enough to prevent condensation of water on the seeds. This is an important consideration, because malted barley is very susceptible to microbial attack; dry conditions must be maintained. Another important consideration is the quality of the barley grains. Plant breeders have developed barley cultivars that are particularly well suited to malting. Such cultivars germinate uniformly, so that at the time of kilning, amylases have been released in all of the seeds and endosperm starch has been partially converted to

Cell Culture and Food

187

simpler carbohydrates. This is important because mashing requires a large number of amylases. Furthermore, partial conversion during malting speeds up the total conversion of starch during mashing. These are not the only reasons for seeking uniform germination. Brewers want all of the seeds to proceed substantially through the germination process because malting results in favorable changes in the flavor of the barley grains. These flavor changes are vital to the quality of many beers. The quality of the seed is also important. It should be clean and free of soil and insects. High grades of malting barley will have a high germination rate (proportion of seeds that germinate), resulting in the accumulation of large amounts of enzyme during malting. If the seed is too old and has a low germination rate, there is unlikely to be sufficient enzyme present during mashing. Old seed is also more likely to be contaminated with fungi that can cause severe problems during malting. During the initial stages of malting, the conditions are ideal for microbial growth, and high levels of contamination of the barley grains usually result in extensive spoilage of the germinating seeds.

C. MASHING Unlike malting, which usually requires dedicated malting operations, mashing typically occurs in the brewery. First, malted barley is ground, releasing starch grains from the seeds. Water is then added to the ground malt, and the mixture is slowly heated to 65°C. Some mashing processes have a prolonged incubation at 40°C to allow proteolysis and protein reduction in the mash. This is necessary if large amounts of protein are present in the barley seed that would result in haze of the finished product if the proteins were left intact. The proteases in barley malt are inactivated by the usual mash temperature (65°C), so slightly prolonged (e.g., 30 min) incubation at a lower temperature is advised if haze is a recurring problem. Eventually, though, the temperature will be increased to 65°C. This is the optimal temperature for β-amylase, which is an exoenzyme that removes maltose (a disaccharide made up of two α-linked glucose molecules) from the nonreducing ends of starch polymers. The action of α-amylase is also important — it is an endoenzyme that attacks starch within the polymer, generating a large number of linear starch chains that are efficiently attacked by β-amylases. Mashing is said to result in total conversion, but this is not quite true. Especially when barley malt is used, a substantial proportion of the starch is converted into short chains of glucose (dextrins) that S. cereviseae is unable to use. The conversion of starch to simpler carbohydrates occurs much more quickly in the mash than in germinating seeds. Within an hour, very little starch remains, and the resulting wort is an excellent medium for yeast fermentation. However, the wort at this stage is not completely ready for yeast inoculation. First, the wort is filtered, either through an artificial filter or through a lauter tun, which is a filtration bed of barley husks and seed coats. The filtered wort is then heated to boiling, for four reasons: • To inactivate amylases and other enzymes. • To precipitate proteins, which will settle out. This will reduce the level of protein haze in the final product.

188

Introduction to Food Biotechnology

OH R

O

Humulone

FIGURE 7.4 General structure of humulones. These compounds provide much of the bitter flavor of hops. Essential oils, which are important aroma compounds, have a similar structure but are smaller and more volatile. The “R” group is a hydrocarbon chain of varying length.

• To allow efficient extraction of resins and essential oils from added hops. • To reduce the level of microbial contamination in the wort.

D. HOPS H. lupulus, the hop plant, is a member of the Cannabinaceae, a family that is renowned for the production of compounds that have strong effects on humans. Cannabis sativa (marijuana) contains secondary compounds that affect the human nervous system. Secondary compounds are, by definition, not involved in primary plant metabolism. Therefore, they are not required for basic processes related to the acquisition of energy and carbon. Why do plants bother to produce secondary compounds? Many secondary compounds are toxic to animals and thus are thought to be an important defense against herbivorous animals. Fortunately for humanity, many secondary compounds are not toxic to humans and are actually highly prized as flavor compounds. Most spices contain flavorful secondary compounds. Hop flowers have many secondary compounds in the form of resins; this includes the humulones and essential oils, which are particularly important in brewing (Figure 7.4). These lipophilic compounds are insoluble in water; they solubilize only through prolonged boiling. They are responsible for much of the bitterness, flavor, and aroma of beer. One of the most crucial jobs of a brewmaster is to add the correct amount of the correct cultivar of hops that will give the desired flavor to a beer. Many of the beer brands produced in North America contain very little hops, because of consumer demand for beer that is not bitter. However, many of the beers produced by microbreweries for niche markets in North America and many European beers contain substantial amounts of hops. In these beers, the type and amount of hops that are added to the wort are extremely important. Some of the resins that are released from hops have antimicrobial properties, particularly against Gram-positive bacteria. Originally, this was probably one of the most important reasons for adding hops to wort. Before the 20th century, brewing

Cell Culture and Food

189

was a risky business; contamination of the yeast inoculum inevitably occurred, and the antimicrobial properties of hops were doubtless essential to the brewer. The antimicrobial properties of hops are less important to modern brewers.

E. PRIMARY FERMENTATION Fermentation is the crucial part of the brewing process. The whole point of brewing is to produce a beverage with ethanol that is pleasing to the senses, and fermentation fulfils that goal. Primary fermentation also has a special place in the history of science. The process of glycolysis and fermentation was first elucidated in S. cereviseae. This was the first biochemical pathway to be deciphered, and it laid a foundation for the development and unfolding of biochemistry in the 20th century. S. cereviseae is still an important model organism, although its role has shifted toward helping unravel the molecular processes that govern cell behavior, especially surrounding cell division. In terms of the technical aspects of fermentation, a few steps are required before adding the yeast inoculum. After boiling, hops are removed from the wort, and the wort is cooled and transferred to a bioreactor. The appropriate lager or yeast strain of S. cereviseae is then added (pitched) to the wort. Hundreds of yeast strains have been isolated that have slightly different growth and fermentation characteristics. The choice of strain sometimes has important effects on the final flavor of the beer. Brewers usually add large amounts of yeast to the wort (e.g., 0.5 L of yeast suspension per hectoliter of wort), so that the yeast does not require a long lag period to grow to heavy populations. As yeast growth begins, so does fermentation. Fermentation is an anaerobic process with the purpose of generating cellular energy through the reduction of organic electron acceptors. This reduction is usually linked to the catabolism of carbohydrates such as glucose (Figure 7.5), although other compounds, such as amino acids, can usually be shunted into fermentation pathways. The fermentation of glucose by yeasts can be summarized by the following equation, which was derived by Gay-Lussac in 1810: Theoretical yield:

C 6 H12 O 6  → 2C 2 H 5OH glucose ethanol 180 g 92 g

+

2CO 2 carbon dioxide 88 g

The wort should be well aerated or oxygenated during the initial stages of fermentation. Although fermentation is an anaerobic process, S. cereviseae requires oxygen for the production of ergosterol, the compound in fungal plasma membranes that confers strength and rigidity. If the initial stages of fermentation are aerobic, the yeasts will be able to synthesize enough ergosterol to supply cellular maintenance needs throughout the fermentation period. If aerobic conditions occur, then why don’t the yeasts use respiration to oxidize carbohydrates? Indeed, it would appear to be foolish for a yeast cell to use the inefficient, low-energy-yielding fermentation pathway in preference to respiration. The solution to this conundrum lies in the relationship between yeast metabolism

190

Introduction to Food Biotechnology

glucose

ATP

ADP glucose-6-phosphate fructose-6-phosphate

ATP ADP

fructose-1,6-diphosphate D-glyceraldehyde-3-phosphate (2) Pi

NAD+ NADH + H+

1,3-diphosphoglycerate (2)

ADP ATP

3-phosphoglycerate (2) phosphoenol pyruvate (2)

ADP ATP

pyruvate (2) CO 2 + NADH + H acetaldehyde (2)

NAD+

ethanol (2)

FIGURE 7.5 Fermentation of ethanol. Glycolysis is followed by conversion of pyruvate into ethanol and carbon dioxide. This process allows regeneration of NAD+ and efficient elimination of waste products of glycolysis (ethanol diffuses freely across membranes). Note that a small amount of adenosine triphosphate (ATP) is gained through fermentation. (NAD, nicotinamide adenine dinucleotide; ADP, adenosine diphosphate; Pi, phosphate.)

and the local concentration of carbohydrates. Many strains of S. cereviseae are unable to respire in the presence of large concentrations of glucose and other carbohydrates. Such conditions are found in wort. Mitochondrial structure is disrupted, and the yeast cells switch to fermentation as the predominant energy-yielding pathway. This effect, known as the Crabtree effect, leads to vigorous fermentation by brewing yeasts despite the presence of oxygen.

F. SECONDARY FERMENTATION Once fermentation is complete and all of the fermentable carbohydrates have been converted to ethanol by the yeasts, the beer is usually transferred to another tank, cooled, and stored. Yeasts are still present at this point. The length of time required for maturation varies from several days to several months. A number of chemical changes occur in the beer; some are a consequence of metabolism by residual yeasts that are unable to grow because of nutrient starvation but are still capable of limited

Cell Culture and Food

191

metabolism. Other changes are strictly due to chemical reactions occurring in the maturing beer. The maturation period is important, because the changes that occur virtually always result in a better-tasting beer. The most important change that occurs during secondary fermentation is the reduction in levels of diacetyl. This compound has a strong buttery taste that is desirable in butter but undesirable in beer. During and after yeast growth, diacetyl forms from α-acetolactate through a nonenzymatic process (α-acetolactate is an intermediate in several amino acid biosynthetic pathways). Yeast can convert diacetyl to acetoin, an unflavored compound, but this conversion is delayed by the slow rate of formation of diacetyl from α-acetolactate. Hence, secondary fermentation sometimes requires several weeks. In many ales, conversion of diacetyl to acetoin is less critical, partly because of the presence of strong flavors that mask the flavor of diacetyl. At the end of the maturation period, the beer is packaged, either in kegs or in bottles — in some traditional operations, secondary fermentation occurs in the kegs or bottles. Modern breweries carbonate the beer before packaging. Fermentable carbohydrates (usually glucose) are added to the kegs or bottles; fermentation by residual yeasts will then result in CO2 production and a carbonated final product.

G. BIOTECHNOLOGICAL IMPROVEMENTS: CATABOLITE REPRESSION Catabolite repression can cause prolonged fermentation times. Wort usually has large amounts of glucose, maltose, and maltotriose. Glucose is a better source of carbon because it can be directly funneled into the fermentation pathway. Consequently, many yeast strains preferentially ferment glucose. Other carbohydrates will be used only when glucose is completely used up (thus use of certain catabolites is repressed by other catabolites). This results in an undesirable decline in the rate of ethanol production. Brewers are acutely aware of the economic costs of delays in the brewing process. When summed over an entire year, small delays can cause significant losses in brewery output. First, we will describe catabolite repression in E. coli, because it is best understood in this organism. In E. coli, and the Enterobacteriacea in general, catabolite repression occurs when glucose is present in the growth medium. High levels of glucose inhibit adenylate cyclase, resulting in low levels of cyclic adenosine monophosphate (cAMP). This depresses expression of genes necessary for catabolism of other carbohydrates (e.g., maltose). However, when glucose is depleted from the medium, adenylate cyclase is no longer inhibited. Levels of cAMP rise, and cAMP interacts with catabolite activator protein (CAP). This interaction allows CAP to bind to promoters upstream from genes and operons that code for enzymes that catabolize carbohydrates. Following this binding of CAP to the promoters, RNA polymerase binds efficiently to the promoters, and transcription can occur. At first glance, catabolite repression seems to be a cumbersome and unnecessary method of gene regulation. However, it is important to carefully consider the natural environment of E. coli — the mammalian gastrointestinal (GI) tract. Nutrient supply in this environment is unpredictable and sporadic. When nutrients appear in the GI tract, intense competition ensues among the resident bacteria. Glucose is an excellent

192

Introduction to Food Biotechnology

source of energy for organisms because it can be directly shunted into glycolysis; other carbohydrates usually have to be enzymatically modified before they can enter glycolysis or other energy-generating pathways. Consequently, E. coli can maintain a higher growth rate when using glucose than when it is using other carbohydrates. In the highly competitive environment of the GI tract subtle changes in growth rate can determine the fate of a bacterium. It makes sense, then, for E. coli to attempt to “grab” glucose when it becomes available and to “ignore” other carbohydrates while glucose is present. Catabolite repression is similarly important for S. cereviseae, which must compete with many other microbes in its natural environment (plant surfaces and sucrose-rich fruits). Because of the importance of yeast metabolism to the brewing and baking industries, a number of researchers have studied catabolite repression (also known as glucose signaling) in yeast. Overall, the phenomenom is similar in yeast as in the Enterobacteriaceae. A pronounced diauxic shift occurs when cells exhaust glucose from the medium. This refers to the process of shutting down genes involved in glucose uptake and utilization and the activation of genes involved in uptake and utilization of other sugars, such as maltose. However, the mechanism of glucose signaling appears to be very different in yeast. cAMP is involved in regulation, but the pattern is different. In E. coli, cAMP levels increase when glucose is depleted, and this triggers transcription of genes involved in utilization of other sugars. However, in yeast, cAMP levels increase when glucose is present in the medium. This shuts down transcription of genes that code for proteins involved in uptake and use of maltose and other sugars. The mechanism of this regulation is unclear, but several complex models have been proposed. Because the molecular mechanisms behind catabolite repression in yeast are still unclear, it is impossible to precisely and directly alter yeast genes in order to modify catabolite repression. Instead, catabolite repression has been eliminated in certain yeast strains using a mutagenesis approach (this approach has been successful with both brewing and baking strains of S. cereviseae) (Figure 7.6). Yeast cells are exposed to a mutagen (a compound that causes random changes in the DNA sequence) and then to 2-deoxyglucose, a synthetic analog of glucose. 2-Deoxyglucose cannot be used as a source of energy and carbon by yeast cells. Cells with normal “wild-type” catabolite repression are unable to use alternative carbon sources because of the inhibiting effect of 2-deoxyglucose. However, if the biotechnologist is lucky, one of the many mutants that occur after mutagenesis will have impaired catabolite repression, perhaps due to a mutation in a gene coding for a key regulator of gene expression. Such a mutant may be able to use other carbohydrates, despite the presence of 2-deoxyglucose. This strategy has been successful in producing mutant yeast strains that can use glucose, maltose, and maltotriose simultaneously, thus avoiding the lag phase that normally occurs as glucose levels decline.

H. HIGH GRAVITY BREWING One strategy that has been employed to improve fermentation efficiency is to use worts that have high specific gravity. Initial specific gravity of a wort is determined

Cell Culture and Food

193

Expose yeast to mutagen

Random mutations occur throughout the yeast genome

Expose yeasts to 2-deoxyglucose

Catabolite repression prevents utilization of other carbohydrates

Wild yeasts die

Yeasts with impaired catabolite repression can use other carbohydrates

Mutant yeasts survive

FIGURE 7.6 The use of a mutagenesis strategy to isolate yeast strains with impaired catabolite repression.

by the concentration of solutes in the wort. Specific gravity is an important measure because of its direct relationship to the final ethanol concentration of the beer. Worts with higher than normal initial specific gravity result in beer with increased ethanol content. This beer can then be diluted, giving a larger volume of beer per fermentation run. This high-gravity approach is not always successful. If the osmotic potential of the wort is too high, a stuck fermentation may occur, because of inhibition of yeast metabolism. If the concentration of ethanol exceeds the tolerance of the yeast, then fermentation will halt. Consequently, one of the aims of brewing research has been to increase the tolerance of yeast strains to high initial levels of solutes and high levels of ethanol toward the end of primary fermentation. With some strains, simple adjustments in the brewing environment allow successful high-gravity brewing. For example, the combination of increased initial oxygen concentration, increased temperature during fermentation (from 14 to 25°C), and increased density of yeast inoculum leads to successful high-gravity brewing in certain yeast strains. Other studies have demonstrated that establishing high levels of oxygen and free amino nitrogen (a good source of nitrogen for yeast cells) is crucial to success in highgravity and very-high-gravity brewing systems. Another crucial development has been the isolation of yeast strains that tolerate high levels of ethanol (as much as 14%) and that also perform well in the brewing environment. With these technological improvements, high-gravity brewing is increasing in popularity among brewers, because of the substantial cost savings.

194

Introduction to Food Biotechnology Glucose Glycolysis Pyruvate

a-acetolactate

Ethanol

a-ALDC

Acetoin

Diacetyl Valine

FIGURE 7.7 Recombinant approach to reducing diacetyl production by brewing yeast. The gene for α-acetolactate decarboxylase (α-ALDC) is transferred to a brewing strain of S. cereviseae. Diacetyl is normally produced from α-acetolactate. α-ALDC diverts α-acetolactate toward the production of acetoin, dramatically decreasing production of diacetyl. (Modified from Onnela, M.-L., et al., J. Biotechnol., 49, 101, 1996.)

I.

THE β-GLUCAN PROBLEM

β-Glucans are polysaccharides that are present in the cell wall of barley seeds. They are released into the wort during mashing, and, although β-glucanases are also present in the mash, they are usually ineffective, because they lack thermotolerance and are inhibited by the temperatures used for mashing. β-Glucans lead to increased viscosity of the wort, which increases the time required for filtration of the mash, especially when filtration occurs through a lauter tun. One approach to tackle this problem is to modify barley so that it produces a thermotolerant β-glucanase. This cannot be done through conventional breeding, so it is necessary to take a transgenic approach. Unfortunately, it was not possible until relatively recently (1990) to produce transgenic barley, primarily because of difficulties in tissue culture of this plant. In the early 1990s, though, several research groups successfully transformed barley, and transgenic plants with thermotolerant β-glucanases appeared in the literature. One major obstacle has been that only certain barley varieties are amenable to transformation, and these varieties are not necessarily the best varieties for brewing purposes. Consequently, extensive backcrossing will be required to achieve satisfactory cultivars with thermostable β-glucanases.

J.

GETTING RID

OF

DIACETYL

As previously discussed, a major objective of secondary fermentation is to decrease levels of diacetyl below 0.5 mg/L (below this threshold, diacetyl flavor is undetectable). Unfortunately, this sometimes leads to extended incubation in large stainless steel tanks, which is expensive to the brewer. One strategy for accelerating the rate of diacetyl removal is to introduce the enzyme α-acetolactate decarboxylase (αALDC). This enzyme catalyzes the conversion of α-acetolactate to acetoin, thus hampering diacetyl production (Figure 7.7).

Cell Culture and Food

195

α-ALDC is not found in S. cereviseae. However, it is present in a number of bacteria, including Klebsiella terrigena and Enterobacter aerogenes. Several research groups have successfully introduced genes coding for this enzyme into brewing strains of S. cereviseae, and when used to ferment wort, these recombinant strains produce beer with undetectable levels of diacetyl after primary fermentation. Thus secondary fermentation is unnecessary with these yeast strains. A number of other recombinant yeast strains are also available; some have introduced amylase genes, leading to improved utilization of carbohydrates, and some have improved flocculating ability (“super-flocculent yeasts”). These latter strains are useful because they are easier to separate from beer after fermentation because they form thick flocs (clumps) that settle to the bottom of the fermentation tank. The beer can then be removed, leaving the sediment behind. Super-flocculent yeasts also tend to be stickier in general, and thus are more easily immobilized to cellulose beads, wood chips, or other substances. Scientists are very interested in using immobilized yeasts for either primary or secondary fermentation, because theoretically this could shorten substantially the time required for fermentation and make it more energy efficient. Brewers could use immobilized yeasts in a continuous fermentation system rather than the almost universal batch system (see Chapter 8, Section III.B, for a discussion of continuous bioreactor operation).

III. DAIRY BIOTECHNOLOGY A. INTRODUCTION Few members of the public realize that when they are consuming yogurt they are consuming live bacterial biomass, suspended in a mixture of acidified milk mixed with bacterial slime layers. Perhaps it is fortunate the public remains ignorant of this; many people would likely consider the idea of eating bacterial biomass to be somewhat disturbing. Nonetheless, yogurt is an excellent example of the utility of large-scale cell culture in the dairy industry. Yogurt, cheese, and buttermilk cannot be made without the participation of a specific group of Gram-positive eubacteria: the lactic acid bacteria. These bacteria are unusual in that they are relatively acid tolerant — they can survive and grow in acidic environments, some as low as pH 3. Other important characteristics of LAB include the following: • They lack pathogenicity. They have a long history of safe use in foods and do not cause disease, except in rare and unusual circumstances. Some of the LAB can cause spoilage (e.g., Pediococcus), but this results in food deterioration, not danger to human health. • They can ferment lactose and other carbohydrates into lactic acid. When inoculated into milk, LAB produce lactic acid, which lowers the pH of the milk. If enough lactic acid accumulates, casein in the milk coagulates, resulting in a semisolid product. In combination with chymosin (rennin), the resulting curd is quite firm, and when pressed and ripened, the curd develops into cheese (see Chapter 3, Section XI.A, for an explanation of

196

Introduction to Food Biotechnology

milk

inoculate with Lactococcus lactis and Lactococcus cremoris

inoculate with Lactobacillus bulgaricus and Streptococcus thermophilus

press curd to remove whey

Ripen (may involve salting)

package (fresh cheese)

stir fruit or other flavoring into the coagulated milk

package (yogurt)

inoculate with Lactococcus cremoris and Lactococcus lactis subsp. diacetylactis

stop fermentation by cooling to 5°C (prevents coagulation)

package (buttermilk)

package (e.g., cheddar)

FIGURE 7.8 The use of lactic acid bacteria to transform milk into a variety of value-added fermented foods.

the action of chymosin). Products with a variety of consistencies are produced by different species; for example, sour cream, buttermilk, yogurt, feta cheese, and cheddar cheese are produced with help from LAB, but they have extremely different textures and consistencies. Each of these products is made with a specific suite of LAB (Figure 7.8). • Certain LAB can improve the flavor of dairy products. For example, the diacetyl that gives buttermilk its characteristic flavor is produced by Lactococcus cremoris. • Partly because of their pH-lowering fermentation products, and partly due to the common production of bacteriocins (compounds that inhibit other bacteria), growth of LAB in milk extends the shelf life of the milk and increases protection against pathogenic bacteria. As is the case with the brewing industry, the overall strategies used for the production of fermented dairy products have not fundamentally changed from traditional practices. Cheese in particular has a long history of use by humans; it is likely that human consumption of animal milk did not occur to a great extent before cheese making was invented, because of the extremely short shelf life of unrefrigerated milk. Cheese has become immensely popular in many cultures, and cheese

Cell Culture and Food

197

making has evolved into a complex art that can produce cheeses with a diverse range of flavors and textures. Cheese shares with wine and few other foods the distinction of being widely popular among average consumers as well as having cult-like aficionados who elevate it to the highest levels of gastronomy.

B. STARTER CULTURES The major application of large-scale cell culture to dairy biotechnology is in the inoculation of starter cultures of LAB to milk, followed by the transformation of the milk into a variety of products. One of the major trends within the dairy industry is toward the use of defined starter cultures, instead of using undefined mixtures of bacteria. This allows improved consistency; a defined mixture of LAB should give consistent levels of lactic acid production and flavor compounds, leading to decreased variability in the time required for production and in the characteristics of the final product. The use of defined starter cultures requires extensive knowledge of the biological characteristics of individual lactic acid strains. Our understanding in this area still has large gaps, but we can make some comments about specific LAB: • Lactococcus lactis subsp. lactis and Streptococcus thermophilus are homofermentative, meaning that they produce only one end product of fermentation. S. thermophilus is unusual in that it is thermophilic, and thus can withstand higher temperatures (it grows well at 40 to 45°C). Most other bacteria used in starter cultures are mesophilic and require temperatures around 30°C. Lactic acid is the only product of fermentation. These bacteria efficiently acidify milk and are extensively used in starter cultures. L. lactis is usually included in cheese starter cultures, and S. thermophilus is usually used in yogurt starter cultures. S. thermophilus is also used in the production of certain hard cheeses (e.g., Emmenthal and Gruyère), because their production requires a heat treatment (~45°C) that nonthermophilic starters cannot withstand. • Leuconostoc mesenteroides subsp. cremoris is a heterofermenter, meaning that it produces more than one end product of fermentation. It produces lactic acid and various flavor compounds (e.g., acetic acid). It is normally used in combination with homofermenters in starter cultures, although it can be used alone as a starter culture for certain cheeses (e.g., Brie and Camembert). • Lactobacillus delbrueckii subsp. bulgaricus and subsp. lactis are homofermenters, producing lactic acid via fermentation. However, they are also important producers of flavor compounds and are popular additions to starter cultures. L. delbrueckii subsp. bulgaricus is usually part of yogurt starter cultures. • Bifidobacterium spp. are obligate anaerobes found in the GI tracts of many animals. They are heterofermenters and are included in starter cultures because of their production of flavor compounds (e.g., acetic acid) and

198

Introduction to Food Biotechnology

because they are considered to be probiotic (health promoting). The probiotic aspects are linked to the ability of bifidobacteria to colonize the human GI tract. In cheese production, the main purpose of the starter culture is to acidify the milk, thus promoting coagulation of the curd. However, in ripened cheese, the starter organisms also play a role in modifying the texture and flavor of the cheese during ripening. Ripening can occur in many ways, but it often consists of cutting blocks of curd, floating them on brine solutions (this increases the salt concentration within the curd, which discourages the growth of spoilage microbes), and then incubating in a cool, humid environment for weeks or months. During this time, enzymes present in the cheese modify the texture and flavor of the cheese. Because the main source of enzymes is from either starter organisms or microbes that survived milk pasteurization, the nature of the microbial flora in ripened cheese is of utmost importance. Recently, there has been a trend toward the use of defined cultures of bacteria or fungi during the ripening process, rather than relying on starter bacteria or “contaminating” bacteria to affect the desired changes. An increasing trend has been to consider the health-promoting aspects of the bacteria that are used in starter cultures (these are called probiotic bacteria). Most attention has focused on Bifidobacterium spp. and Lactobacillus rhamnosus GG. These bacteria are able to colonize the human intestine and are thought to provide protection against certain intestinal pathogens (e.g., Clostridium difficile). Although they do not contribute greatly to acid or flavor production, they may make the final product more attractive to health-conscious consumers. More conventional starter organisms such as Lactobacillus acidophilus and Lactobacillus bulgaricus also appear to have probiotic potential, although the benefit to the consumer may lie in modulation of the immune system.

C. PHAGE The biggest problem encountered during the fermentation of milk is starter failure due to bacteriophage. Phage can spread rapidly through a fermentation vat and completely stop the acidification of milk, leaving the milk vulnerable to spoilage by competing microorganisms. Pathogenic organisms such as Staphylococcus aureus may also grow to high densities, creating a potential public health problem through the accumulation of toxins. The source of phage is either the milk or the starter culture. Many phage are heat resistant and are unaffected by pasteurization; for this reason, fermentation processes may need to include a more rigorous heat treatment of the milk (90°C for 30 min). In most cases, though, the emphasis is on prevention of phage infection of the starter cultures. This is usually accomplished by using phage-inhibitory media (see below) to grow starter organisms, if the cheese maker produces its own inocula of starter cultures for its milk fermentations. A popular alternative is to purchase freeze-dried cultures that can be directly inoculated into the fermentation vat, thus avoiding potential phage contamination during in-house starter cultivation.

Cell Culture and Food

199

Fortunately, phage infection does not automatically lead to starter failure. Like many other viruses, there is strain-to-strain variation in virulence. Virulent phage proceed through the lytic cycle (see Chapter 2, Section IV.A), killing the host bacterium, whereas temperate phage have a period of dormancy that does not immediately lead to death of the host cell. Some strains exhibit intermediate levels of virulence and only cause starter failure if fermentation vats are not completely sanitized between production runs. Such phage often replicate to high numbers (e.g., 108 per milliliter) without significantly affecting the growth of LAB or, more importantly, the rate of lactic acid production. However, if this level of inoculum is still present when a new batch of milk is added, then starter failure is likely. Lysogenic strains can also cause starter failure, because mutant, virulent viruses occasionally arise from such strains. Also, most lysogenic strains have a low frequency of spontaneous induction of lytic phage; consequently, most cultures that carry lysogenic phage also have “free” phage that have been released via the death of phage-containing cells that have been induced into the lytic cycle. In many countries (e.g., France) artisan cultures, used by small-scale manufacturers of specialty cheeses, are a potent reservoir of lysogenic phage and can lead to dispersal of phage to other cheese makers. Thus, phage is a prevalent and potentially devastating problem for cheese makers and other manufacturers of fermented milk products. Fortunately, two strategies are available to combat phage: • Phage-resistant strains (also known as bacteriophage-insensitive mutants, or BIMs). LAB are not universally susceptible to phage. Genetically derived resistance in some cases is due to well-characterized restriction–modification systems (the combination of restriction enzymes that degrade phage DNA and methylases that protect bacterial DNA from the restriction enzymes — see also Chapter 3, Section IV.A). In other cases, the resistance is less well understood, but it appears to be related to bacterial interference with phage absorption or injection. Many of the resistance genes are plasmid encoded and therefore can be transferred to other LAB. • The use of phage-inhibitory media to cultivate starter organisms. It has been known since the 1960s that the addition of calcium chelators to milk inhibits phage multiplication. Calcium is required for phage multiplication, and the addition of chelators results in suppression of a wide range of phage. The only drawback to this system is that the chelators often inhibit the growth of LAB as well.

D. RECOMBINANT LACTIC ACID BACTERIA 1. Improved Starters Researchers have shown great interest in the last decade in developing techniques to transform Lactococcus spp., Lactobacillus spp., and other LAB. This is partly

200

Introduction to Food Biotechnology

due to their importance to the dairy industry, but also because of their importance to other food and beverage fermentations, such as the use of Lactobacillus plantarum in sauerkraut production and the use of Leuconostoc spp. to drive the malolactic fermentation in wine. In the latter process malic acid is converted to lactic acid, decreasing the sour taste of certain wines. Unfortunately, the LAB are more difficult to transform than bacteria such as E. coli. However, considerable progress has been made, and successful reports of recombinant LAB are appearing with greater frequency in the literature. As an example, we will discuss the production of stable recombinants of Lactobacillus casei that have immunity to phage A2. Plasmid vectors are available for the transformation of LAB, but plasmids have strong disadvantages for use as cloning vectors for LAB utilized in food fermentations. The main reason is that plasmids are unstable (can be lost by the host bacterium) in the absence of selective pressure that favors plasmid retention. In routine cloning experiments with plasmid vectors, this problem is solved by inserting antibiotic-resistant genes into the vector. Antibiotics in the growth media then exert selective pressure for maintenance of plasmids within host cells (see Chapter 3, Section IV.B). Obviously, it is impractical (and illegal in most countries) in a food fermentation setting to add large amounts of antibiotics. There is also a less obvious problem — antibiotic-resistant genes are considered to be non-food-grade DNA. The incorporation of this DNA would mean that the LAB would lose its GRAS (generally recognized as safe) status and would not be usable in food fermentations. LAB used in food fermentation have GRAS status, which means that the U.S. Food and Drug Administration (FDA) does not consider these bacteria to be food additives. Their use in food is less restricted than would be the case if they were considered to be additives. The best way to circumvent the problem of plasmid instability is to integrate desired DNA sequences into the bacterial genome. A Spanish research group recently demonstrated that this is feasible, by inserting a gene from phage A2 into the chromosome of L. casei. The resulting recombinant strain is immune to infection by phage A2. Two plasmid vectors were used (Figure 7.9). One plasmid (pEM76:cI) contained the phage gene (cI) as well as a gene (int) that directs integration of associated DNA into a specific region of the L. casei genome. DNA sequences designated as six were also present in this vector; they flanked a sequence of DNA containing several antiobiotic-resistant genes. The second plasmid (pEM68) contained a gene coding for β-recombinase. This gene codes for an enzyme that deletes DNA between six sequences. L. casei was first transformed by electroporation with pEM76:cI (Figure 7.10). The integrase coded by int then catalyzed insertion of this plasmid into the chromosome of L. casei. The resulting transformants could be selected because of their antibiotic resistance. Note that pEM76:cI does not have an ori for L. casei; hence, cells would grow in the presence of the antibiotic only if pEM76:cI was integrated into the genome. The next step was to remove the antibiotic-resistant gene once recombinant bacteria were isolated. This was achieved by transforming these bacteria with pEM68. The β-recombinase then directed the removal of the antibiotic-resistant genes because they were flanked by six sequences. This led to bacteria that were

Cell Culture and Food

201

ampr six

six

b-recombinase

int

cI

pEM76:cI

pEM68

FIGURE 7.9 Plasmid vectors used to transform Lactobacillus casei with stable phage resistance. (ampr, ampicillin resistance.) (Data from Martin, M. C., et al., Appl. Environ. Microbiol., 66, 2599, 2000.)

sensitive to the antibiotics but had the phage A2 gene stably integrated into the bacterial genome. When challenged with phage A2 in milk fermentation assays, the recombinant bacteria were immune. This occurred because virally infected cells are often immune to infection by other viruses. In this case the presence of a single viral gene was sufficient to generate a similar degree of immunity. 2. Recombinant Lactic Acid Bacteria as Vaccines LAB have great potential to deliver vaccines that protect against human disease. Many have GRAS status, making them amenable to oral delivery of vaccines. Oral vaccines are preferable to injected vaccines because they are better suited to distribution and use, particularly in the developing world. They also tend to stimulate the mucosal branch of the immune system. Because most pathogens enter the human body by breaching mucosal defenses (e.g., intestinal or lung mucosal epithelia), mucosal immunity is considered to be more protective than a more generalized immune response. Food-grade LAB are also attractive as vaccine delivery systems because they are closely related to several pathogens, including Streptococcus pyogenes and Streptococcus pneumoniae. Thus, they have a similar genetic background to these pathogens. The importance of this similarity is illustrated by a project that involved transfer of S. pneumoniae genes required for capsule production to L. lactis. These capsules form a polysaccharide covering for S. pneumoniae and are key virulence factors (traits related to the ability of a pathogen to invade and cause damage in a host), because they help the bacterium evade engulfment by phagocytic cells of the immune system. The genes were successfully inserted into L. lactis via a plasmid vector, and the resulting recombinant bacteria, after injection into the peritoneal cavity of mice, provoked an antibody response against S. pneumoniae capsules. When a similar strategy was attempted with E. coli, intact capsules were not assembled and thus were incapable of causing the desired immune response. Evidently,

202

Introduction to Food Biotechnology Transform L. casei with pEM76:cI pEM76:cI Chromosome

Plasmid is inserted into the chromosome

Plate on ampicillin nonrecombinant cells die

Transform with pEM68

amp r is removed from chromosome

FIGURE 7.10 Method used by Martin et al., to transform Lactobacillus casei with stable phage resistance. (Data from Martin, M. C., et al., Appl. Environ. Microbiol., 66, 2599, 2000.)

the genetic and cytoplasmic environment of L. lactis was more similar to that of S. pneumoniae, allowing correct assembly of the polysaccharide capsule. We have discussed only two of many potential examples of the utility of recombinant LAB. The broad use of these transformants, though, will depend on further research into technical aspects of transformation and expression of heterologous proteins in LAB and on public acceptance of products derived from recombinant LAB.

IV. AMINO ACIDS: NUTRITIONAL BOOSTS AND FLAVOR ENHANCERS A. OVERVIEW Amino acid harvesting from microbial cultures is an economically important industry. The Japan Amino Acid Association estimates that the worldwide production of amino acids is at least 1.66 million tons per year. Why produce such vast amounts of

Cell Culture and Food

203

amino acids when they are part of the normal human diet and can be obtained from an incredible variety of animal and plant sources? The answer lies in six main areas: 1. Flavor enhancers. Monosodium glutamate is a flavor enhancer that is added to many processed foods (e.g., dehydrated soups). Glutamate production is estimated to be greater than 1 million tons per year. 2. Supplements. Many animal feeds are low in specific amino acids such as lysine and methionine. The addition of purified amino acids from microbial sources is a relatively inexpensive way of improving the nutritional quality of feed. Worldwide production of lysine and methionine is estimated to be at least 800,000 tons per year. Humans also consume amino acid supplements; from a nutritional perspective, this is unnecessary for people that ingest protein from a variety of sources. However, some alternative health practitioners recommend consumption of amino acids for a variety of ailments. Indeed, certain amino acids have pharmacological effects, because of their structural similarity to human metabolites (e.g., tryptophan and serotonin). 3. Sweeteners. Aspartame is a popular ingredient of soft drinks and many processed foods. It is a methyl ester of aspartyl-phenylamanine, which is made using aspartate and phenylalanine. 4. Parenteral nutrition (e.g., intravenous infusions). Although this is less important economically, the development of technology for the production of many amino acids has been spurred by the need for nutritive, injectable solutions for people who are unable to ingest food. 5. Industrial uses. Certain amino acids have nonfood uses. For example, phenylalanine is used as a component of detergents, and as a chelator during water purification. 6. Pharmaceuticals. Amino acids such as L-proline and D-alanine are used as building blocks for the construction of pharmaceuticals. These applications have led to the establishment of industrial operations that produce billions of dollars worth of amino acids every year. These industries did not develop overnight, though. With the exception of glutamate, it has proven to be extremely difficult to find microorganisms that naturally overproduce amino acids. Consequently, much effort has been expended to modify microbes in order to achieve overproduction. The most successful method so far has been to isolate mutants that have defective control over amino acid production. A number of recombinant DNA approaches have also been successful. The latter process is an example of metabolic engineering and is only possible if we understand the pathways of amino acid biosynthesis and related pathways (e.g., those supplying required coenzymes and substrates). The relationships between biosynthetic pathways and other pathways are also important, because metabolic engineering toward amino acid overproduction will inevitably disrupt overall metabolic flow. This can be visualized in terms of overall carbon and nitrogen; if substantial amounts of carbon- and nitrogen-containing compounds are diverted toward the production of a specific amino acid, that will reduce the flow of carbon and nitrogen to other pathways. Successful metabolic engineering requires a complete understanding of the ramifications of such disruptions.

204

Introduction to Food Biotechnology

COOH NH2

C*

R L-amino acid

COOH H

H

C*

NH2

R D-amino acid

FIGURE 7.11 Stereochemistry of amino acids is dictated by the position of the chiral carbon (*) and its relationship with the nitrogen in the amino group. Solid lines indicate bonds that project out of the page, and dashed lines project in the opposite direction. D- and L-amino acids are present in many organisms (e.g., carnosine in the muscle tissue of cows and other vertebrates), but only L-amino acids are constituents of proteins.

In many cases, we do not have a sufficient understanding of amino acid metabolism to precisely engineer amino acid overproduction via recombinant DNA technology. This is why mutagenesis approaches are still popular, because they can be applied to “black box” systems that are incompletely understood. This section examines the use of both mutagenesis (proline) and recombinant DNA approaches to achieve amino acid overproduction. We will also discuss examples of the use of natural amino acid overproducers (Corynebacterium glutamicum and glutamate) and the use of immobilized enzymes to produce amino acids (aspartate). One question frequently arises: Why don’t we obtain amino acids from meat or other protein-rich foods? In fact, this is a valid approach that is sometimes successful. The major problem is that natural protein sources contain a wide range of amino acids; separation of a specific amino acid from all the others is expensive. Another option is to produce the amino acid synthetically using the many elegant methods developed by organic chemists. The major problem with this strategy is that it is difficult to synthesize specific stereoisomers of an amino acid. Amino acids exist as D and L stereoisomers (Figure 7.11); D-amino acids are present in all cells but not as a constituent of protein. For many of the applications of amino acids, only one of the isomers is useful. The reason for this is clear in the case of amino acids used as feed supplements or infusions — if the ultimate aim is to supplement a human or animal with amino acids that will be used to build protein, then it is essential to use L-amino acids. In other applications, the stereochemistry is equally important. For example, L-glutamate is an effective flavor enhancer, but D-glutamate is completely ineffective. For these reasons, it is usually preferable to use microbes or microbial enzymes to produce most amino acids, so that rigid control over the optical characteristics of the amino acid can be achieved.

B. CHOICE

OF

MICROBE

It is easy to convince microbes to produce large amounts of an end product of fermentation (e.g., ethanol). Simply supply sufficient fermentable substrates and the correct environment. In contrast, the overproduction of amino acids is a challenging

Cell Culture and Food

205

a-ketoglutarate (diverted from the TCA cycle )

transaminase

glutamate ATP glutamate kinase ADP glutamic g-semialdehyde NADH glutamic g-semialdehyde dehydrogenase NAD+

D1-pyrroline 5-carboxylate NADH

D1-pyrroline 5-carboxylate reductase NAD+ L-proline

FIGURE 7.12 Production of glutamate and proline by Serratia marcescens and the associated energy costs. The cell must invest energy in the form of ATP, NADH, and α-ketoglutarate. Energy must also be expended to synthesize four enzymes. Finally, energy is lost because of the diversion of α-ketoglutarate from the TCA cycle. The dashed arrow indicates feedback inhibition by proline.

proposition. Fermentation end products normally leave cells quickly, either through diffusion across membranes (e.g., ethanol) or through active transport (e.g., lactic acid). Thus it is not difficult to separate product from cells, if desired. However, microbes do not normally secrete amino acids, because they are usually required within the cell. Secretion of amino acids is energetically unwise for microbes, because amino acid synthesis typically requires the coordinated activity of several enzymes (Figure 7.12). This makes amino acid production an expensive process that is necessary, but must be tightly controlled, so that costly, unnecessary overproduction does not occur. The secretion of amino acids serves no purpose for a microbe, and therefore it is not surprising that the isolation of a bacterium that naturally overproduces an amino acid is extremely rare. Consequently, biotechnologists have been forced to artificially manipulate amino acid metabolism to achieve overproduction.

206

Introduction to Food Biotechnology

The choice of organism is critical to the development of an overproducing microbe. E. coli would seem to be an ideal microbe for the development of an overproducing strain because it is such a well-characterized organism. However, E. coli is unsuited for most amino acid applications, with some important exceptions (e.g., aspartate). Amino acid production in E. coli is controlled through regulation of enzyme activity, transcription of mRNA, and translation. First, we need to understand how this is achieved: 1. Regulation of enzyme activity through feedback inhibition. Many enzymes have an allosteric site, which is spatially separated from the active site. The end product of the pathway specifically binds to the allosteric site, which results in a conformational change that reversibly inactivates the enzyme. The affected enzyme is typically at the beginning of the pathway. This allows the pathway to be shut off, without risking the build-up of intermediates of the pathway. In the case of proline synthesis (see Figure 7.12), proline allosterically inhibits glutamate kinase. 2. Regulation of transcription (e.g., repression). This is conceptually similar to feedback inhibition; the difference is that enzyme activity is unaffected, but transcription of enzymes is repressed (Figure 7.13). The end product of an amino acid synthetic pathway typically binds reversibly to a co-repressor protein. This binding results in an active repressor complex, which binds to the operator region of the appropriate operon. In bacteria, the genes for pathways of amino acid synthesis often follow a single promoter, so that all enzymes of the pathway are synthesized together. This type of gene arrangement is an operon. As long as the corepressor is bound to the operator, none of the genes of the operon will be expressed, and none of the enzymes will be synthesized. When cytoplasmic levels of the amino acid fall, the repressor complex disassociates, and the promoter can then interact with RNA polymerase to initiate transcription. 3. Attenuation. This complex mode of regulation acts during translation. If a particular amino acid is present at high concentrations in the cytoplasm, then transfer RNA (tRNA) charged with that amino acid will also exist at high levels. For some amino acids, this leads to specific disruption of translation of the operon responsible for synthesis of that amino acid. In E. coli, overproduction of an amino acid can occur only if all of these levels of regulation are overcome. This is possible and has been successfully accomplished in some cases, but, in most cases, better results have been obtained using other bacteria. Serratia marcescens is much easier than E. coli to convert to an amino acid overproducer. It is a member of the Enterobacteriaceae family, and its major claim to fame is its distinct red pigmentation. In this bacterium, feedback inhibition is the most important mechanism of regulation of amino acid synthesis. Why does E. coli have such complex mechanisms of regulation, but for S. marcescens simple regulation of enzyme activity is sufficient? The answer lies in the very different environments of

Cell Culture and Food

207

En z

ym

e

2 En z

ym

e

1 e ym En z

or

er

er at

ot

op

m Pr o

3

A. Proline is present

Proline + Co-repressor

RNA polymerase

Transcription is blocked

B. Low levels of proline

Transcription is unblocked

FIGURE 7.13 Regulation of amino acid synthesis by co-repression. (A) When the amino acid is present in the cytoplasm, it combines with a co-repressor protein to form a complex that binds to the operator site of the operon. That site contains genes coding for enzymes that are responsible for synthesis of the amino acid. When the complex is bound to the operator, RNA polymerase is unable to transcribe the genes of the operon. (B) When the amino acid is present at low levels, the co-repressor alone is unable to block transcription, which proceeds through the operon.

these bacteria. E. coli lives in the mammalian GI tract, an environment of extreme fluctuations in amino acid supply. Bacteria are unable to predict when their hosts will eat, so they must have the ability to rapidly adjust their rates of amino acid synthesis. As an example, consider a human who eats a bowl of of pasta primavera. The food contains an abundance of carbohydrates, but relatively few proteins and amino acids. If E. coli is to grow and compete with other bacteria, it must rapidly assimilate carbohydrates and synthesize amino acids from these carbohydrates. Rapid derepression of anabolic pathways is therefore essential. However, if the human host then consumes several glasses of milk, a rich source of amino acids,

208

Introduction to Food Biotechnology

then the bacterium must adapt quickly. Amino acid synthesis is no longer necessary and is energetically wasteful, because it diverts energy from other pathways. Hence, E. coli requires fine-tuned regulation of amino acid biosynthetic pathways. In contrast, S. marcescens lives in the soil, an environment that is usually starved of free amino acids. Unless the bacterium is lucky enough to find itself in a proteinrich habitat (e.g., a decomposing racoon), it must synthesize its own amino acids. The rate of synthesis must be metabolically tied to the rate of growth, but this can be done quite simply, by feedback inhibition. If the cell is not growing, it has few needs for amino acids; consequently, amino acids accumulate in the cytoplasm and inhibit pathways leading to their synthesis. However, if the cell is growing, it will use up its store of amino acids, and inhibition of anabolic pathways will no longer occur.

C. PROLINE: MUTAGENESIS LEADS

TO

OVERPRODUCTION

Proline is synthesized on an industrial scale for use in intravenous infusions. It is produced through a relatively simple pathway that uses glutamate as the initial substrate (see Figure 7.12). In S. marcescens, proline biosynthesis is regulated predominantly by feedback inhibition. In developing a strain of S. marcescens that overproduces proline, the first obstacle was overcoming feedback inhibition. This was accomplished using a similar approach to that used to overcome catabolite repression in brewing yeasts (this chapter, Section II.G). Mutants were plated onto media containing proline analogs. Wild-type cells died because they were unable to synthesize proline. This occurred because the analogs shut down proline biosynthesis through feedback inhibition. However, mutants that had impaired feedback inhibition were able to synthesize needed proline and were able to survive and grow. These mutants overproduced proline, but further increases in the rate of proline synthesis were required. The following steps resulted in strains of S. marcescens that dramatically overproduced proline: 1. The gene for an enzyme that degrades proline was deleted from the chromosome of S. marcescens. This enzyme allows the bacterium to scavenge proline from its environment and use it as a carbon and nitrogen source, but is unnecessary and undesirable in a proline overproducer. 2. It was observed that growth of S. marcescens in a high-salt medium results in further increases in the rate of proline production. Serratia, like many other bacteria, uses proline as an osmoprotectant when growing in osmotically stressed environments. Osmoprotectants are compounds that can accumulate within cells to high concentration without adversely affecting cellular metabolism. This is beneficial in high-salt (low-water activity) environments because it decreases cellular water activity enough to allow water to flow into the cell through osmosis. This prevents cell dehydration. When grown in a high-salt medium, the bacteria produced 60 to 75 g/L of proline, and ~20% of the carbon flow was directed to proline production.

Cell Culture and Food

209

3. Finally, through recombinant DNA techniques, the genes coding for enzymes required for proline synthesis were cloned and inserted into plasmids that maintain a high copy number in serratia. This resulted in a further 50% increase in the rate of proline production. Proline overproduction, then, was possible through the use of a random process (mutagenesis) combined with exploitation of an aspect of microbial stress physiology (osmoprotectants) and the use of recombinant DNA techniques. Although overproducing strains were achieved, it is clear that the same strategy would not work with other amino acids. One reason for this is that other amino acids are not used as osmoprotectants by microbes, so growth in high-salt media would not result in increased amino acid production. Also, some amino acids (e.g., leucine) are produced by branched pathways, in which several different end products exert feedback inhibition. This makes it much more difficult to obtain mutants deficient in feedback inhibition. Other strategies must be implemented to obtain overproducers of these amino acids.

D. GLUTAMATE: NATURAL OVERPRODUCTION Glutamate is an interesting amino acid for several reasons. It was the first amino acid to be produced from microbial culture on an industrial scale, and, along with L-alanine and L-valine, is the only amino acid that is overproduced by wild (unmodified by humans) bacterial strains. Furthermore, glutamate is economically the most important amino acid; it is extensively used within the food industry as a flavor enhancer and is a popular flavor ingredient throughout the world. Glutamate holds a special place in sensory science because it is the fifth basic taste. Human taste buds have specific receptors for glutamate, along with receptors for sour, sweet, bitter, and salty tastes. The taste of glutamate (umami) is usually described as “brothy.” The flavor-enhancing capabilities of glutamate were utilized in Japan prior to the availability of purified monosodium glutamate. Dried kelp (konbu) has a long history of use as a food seasoning in Japan, and in the early 20th century K. Ikeda, a Japanese scientist, discovered that glutamate was responsible for the flavoring effects of konbu. Then, in the 1950s, came an unusual finding. Bacteria that naturally overproduce glutamate were isolated from the soil. They belonged to the corynebacterium–brevibacterium group and are still used today to produce glutamate on an industrial scale. Although the taxonomy of these bacteria is still somewhat unclear, isolates used for overproduction of glutamate are usually called C. glutamicum. This isolation of a natural overproducer was fortuitous. As explained previously, microbes do not normally overproduce amino acids, unless they are used as osmoticants in media with a high solute or low water content. C. glutamicum does not use glutamate as an osmoticant, and, in any case, the bacteria secrete large amounts of glutamate, an action that would not be taken with an osmoticant (to protect a cell from the adverse effects of low water potential, osmoticants must be retained within the cell). What other reasons could a microbe have for the overproduction of an amino acid?

210

Introduction to Food Biotechnology

After much research, this question has been partially answered. We now know that glutamate overproduction in C. glutamicum occurs for two reasons: (1) under certain conditions (e.g., biotin starvation) the membranes become highly permeable to glutamate, and (2) under conditions of low availability of oxygen catabolic pathways shift toward the overproduction of glutamate. The combination of these two phenomena results in the secretion of large amounts (100 g/L) of glutamate. We do not completely understand why the membranes of these bacteria become selectively permeable to glutamate. It was previously thought that this permeability was a direct result of biotin starvation, because addition of biotin to the growth medium prevents the secretion of glutamate. Biotin is required for the synthesis of fatty acids, so biotin deficiency would be expected to interfere with the synthesis of membranes. However, such effects on membrane structure and integrity should have a wide range of effects on membrane permeability, and should not specifically change the permeability to glutamate. The answer to this puzzle appears to be related, at least partly, to the presence of specific transporters in the membrane of C. glutamicum that pump glutamate out of the cell. Perhaps these pumps are specifically activated under conditions of biotin starvation. The shift in catabolism under conditions of low-oxygen partial pressure is also poorly understood. The respiratory pathway of C. glutamicum is unusual. Under aerobic conditions, carbohydrates are completely respired through a glyoxylate shunt (Figure 7.14), which is a modification of the conventional tricarboxylic acid (TCA) cycle. Under anaerobic conditions, the bacteria switch to fermentation, but in conditions of low-oxygen partial pressure, fermentation does not occur. The glyoxylate shunt also operates at low levels, probably because there is insufficient oxygen to oxidize the reduced nucleotides (e.g., NADH) that are normally generated by the glyoxylate shunt. The bulk of carbon flow from acetyl-CoA is toward α-ketoglutarate, and then to glutamate, which is secreted from the cell. In a normal TCA cycle, α-ketoglutarate is converted into succinyl-coA, generating NADH. The α-ketoglutarate dehydrogenase [α-KDH]) has enzyme that catalyzes this reaction (α very low activity in C. glutamicum. The cell rids itself of α-ketoglutarate by converting it to glutamate and actively secreting it from the cell. The presence of specific pumps for glutamate may be a clue that C. glutamicum derives an energetic benefit from the secretion of glutamate. Perhaps under conditions of low oxygen availability, when there is insufficient oxygen to fully oxidize carbohydrates to CO2, there is still enough oxygen to oxidize a small number of NADH molecules. When glutamate is being secreted, NADH is produced through the formation of acetyl-CoA from pyruvate. Oxidation of this NADH by the electron transport chain would presumably give the cell more adenosine triphosphate (ATP) than it would derive from fermentation. Thus, glutamate secretion may be energetically wise for C. glutamicum. Some gaps still exist in our understanding of glutamate secretion in these bacteria. Fortunately, this does prevent us from growing C. glutamicum on an industrial scale for the production of glutamate. We are fortunate that the environmental conditions that lead to glutamate overproduction and secretion (i.e., low levels of biotin and oxygen) are easily achieved. Growth media such as molasses contain biotin, but it tends to be used up fairly quickly. Oxygen levels also decline quickly

Cell Culture and Food

211

A. Catabolism of pyruvate in the presence of oxygen pyruvate acetyl-CoA

citrate

oxaloacetate malate

isocitrate glyoxylate

glyoxylate shunt B. Catabolism of pyruvate under low oxygen availability pyruvate acetyl-CoA

citrate isocitrate

secreted from cell

α-Ketoglutarate

glutamate

FIGURE 7.14 Carbohydrate catabolism in Corynebacterium glutamicum. (A) Under aerobic conditions, C. glutamicum operates a glyoxylate shunt, which is a truncated form of the TCA. (B) Under conditions of low oxygen availability, it diverts carbohydrates to glutamate production and secretion.

in dense microbial cultures. Consequently, normal methods of growing microbes lead to the correct environmental conditions for overproduction of glutamate.

E. ASPARTATE: CONVERSION

BY IMMOBILIZED

ENZYMES

Aspartame, the methyl ester of aspartyl–phenylalanine, is a popular alternative sweetener for soft drinks and a range of products aimed at the calorie-conscious market. Industrially, aspartame is produced through artificial (i.e., without the use of enzymes or cells) synthesis using aspartate and phenylalanine as reactants. Both of these amino acids are industrially produced using bacteria, although in very different ways. Phenylalanine is usually produced by large-scale growth of specific strains of soil bacteria (either S. marcescens or C. glutamicum). These strains are obtained using a similar strategy as described for obtaining overproducers of proline (i.e., mutagenize and then isolate mutants that have defective feedback inhibition). Phenylalanine can also be produced using immobilized enzymes derived from E. coli, but this is not economically feasible because of the high cost of the substrate (trans-cinnamic acid). Although not feasible for phenylalanine, the use of immobilized enzymes is an economic and popular route for the production of aspartate. The enzyme used in this process is aspartase, the enzyme that catalyzes the degradation of aspartate

212

Introduction to Food Biotechnology

into ammonia and fumarate. Aspartase is normally used by E. coli to scavenge aspartate from the external environment as a source of carbon or energy. However, it will efficiently convert fumarate to aspartate if the concentration of fumarate and ammonia are high (2 M). Most chemical reactions can proceed in both directions, and the tendency for a reaction to proceed in either direction is affected by the equilibrium constant (Km) and the concentrations of the reactants and products. In the case of aspartase, if the initial concentration of aspartate is low, and fumarate and ammonia are present at high concentration, the reaction will proceed toward almost total conversion of fumarate to aspartate. Purified aspartase can be immobilized onto solid supports, but it is equally efficient, and technically easier, to immobilize intact cells of E. coli. Carageenan works well as a solid support; cells of E. coli are added to molten carageenan, which is then cooled, granulated, and packed into a column. Usually, autolysis of the cells occurs within several days of immobilization. The enzyme is released from cells, but is entrapped within the carageenan, and remains stable for several years. Concentrated fumarate and ammonia are added to the top of the column, and aspartate is collected from the bottom. This process is ideal for the large-scale manufacture of aspartate, for three reasons: 1. The substrates (fumarate and ammonia) are relatively inexpensive. 2. Aspartate is produced in a concentrated form, making purification easy. 3. Complex growth media are not required; therefore, fewer unwanted solutes are present in the effluent from the column, simplifying purification. Enzyme immobilization is also used industrially to produce L-alanine, L-cysteine, and a number of other amino acids and amino acid derivatives. D-p-hydroxyphenyl-glycine, L-dihydroxy-phenylalanine,

F. TRYPTOPHAN

AND

EOSINOPHILIA–MYALGIA SYNDROME

Because amino acids have specific industrial uses, it is crucial that amino acid production lines end with efficient and effective purification. Contamination of an amino acid by another amino acid or other cellular metabolites would likely interfere with the end use of the amino acid. In the 1980s, an epidemic of eosinophilia–myalgia syndrome illustrated tragically that amino acid purification is also crucial to consumer safety. Between 1988 and 1990, 25 people, mostly in Japan, died from eosinophilia–myalgia. Hundreds of other people were injured, and the cause of illness was determined to be the consumption of tryptophan supplements. It was subsequently discovered that the tryptophan consumed by these people came from one company and contained trace levels of several potentially toxic compounds. After a thorough investigation, the presence of these toxins was traced to three factors: 1. A strain of Bacillus amyloliquefaciens was used to produce the tryptophan. This strain had recently been modified using recombinant DNA techniques, resulting in overproduction of the product of the first reaction in the synthetic pathway — phosphoribosylpyrophosphate.

Cell Culture and Food

213

2. A step in the purification process was changed. Some of the tryptophan bypassed a reverse osmosis membrane filtration step (see Chapter 8, Section IV.C, for an explanation of reverse osmosis). 3. Lower amounts of activated charcoal were used to remove impurities in the tryptophan. If only one of the preceding factors had been operative, the product may have been safe. However, the combination of increased levels of tryptophan and reduced purification efficiency had lethal consequences. The most important lesson learned from this tragedy is that the overproduction of amino acids can potentially lead to the accumulation of toxins; this is probably related to the reactivity of amino acids, particularly when they are present at high concentrations. Purification is an expensive component of amino acid production systems. Consequently, there is a powerful temptation for companies to cut corners and allow less-purified products to be sold. The eosinophilia–myalgia scandal amply illustrated that this is an unwise course of action. Many opponents of recombinant DNA technology use eosinophilia–myalgia as an example of the potential disasters caused by the production of genetically engineered organisms. This is misleading, because this particular disaster could easily have happened if nonrecombinant methods (e.g., mutagenesis) had been used to boost production of tryptophan. Also, it is unclear which of the three causative factors were most important to the appearance and persistence of the toxins. Eosinophilia–myalgia is a potent reminder of the need for regulatory agencies to monitor biotechnology and new methods of producing biotechnological products, regardless of whether recombinant DNA technology is involved.

V. MICROBIAL ENZYMES A. OVERVIEW Enzymes have a rich history of use in the food industry. They are used in the production or processing of starch, flour, cheese, fruit juices, artificial sweeteners, and meat, and are also frequently used in brewing and wine making. Enzymes are also used in agriculture; for example, dairy farmers use silage (corn and other crops fermented by LAB) as a nutritive, easily stored feed for cows. Fermentation by LAB lowers the pH of the silage, which inhibits the growth of spoilage organisms, as long as air is eliminated from the silage. Cellulases are often added to silage to increase the amount of fermentable sugars. Within the food industry, enzymes are particularly useful if a specific chemical transformation is required. The transformation can be a hydrolytic breakdown (e.g., digestion of proteins by proteases), an additive reaction (e.g., the use of aspartase to synthesize aspartate), or a rearrangement (e.g., conversion of glucose to fructose by glucose isomerase). Usually enzymes are used in their “forward” direction (e.g., proteases are used to digest proteins); however, they are sometimes useful in their reverse direction. For example, we saw in the preceeding section of this chapter that aspartame is made from L-phenylalanine and L-aspartate. The industrial process of

214

Introduction to Food Biotechnology

TABLE 7.2 Applications of Enzymes to Food Production and Processing Enzyme

Use

Product

Source of Enzyme

α-Amylase

Starch processing

Dextrins

β-Amylase Glucoamylase

Brewing (mashing) Starch processing and brewing Fructose production Candy processing Starch processing Juice clarification Milk coagulation Milk coagulation Brewing (mash) Cheese making Dairy processing

Maltose Glucose

Bacillus subtilis Bacillus licheniformis B. subtilis Aspergillus niger

Fructose Glucose + fructose Debranched starch Galacturonate Cheese curd Cheese curd β-Glucose Flavor compounds Glucose + galactose

Streptomyces spp. Saccharomyces cereviseae Klebsiella Aspergillus oryzae Kluyveromyces spp. Mucor miehei A. niger Rhizopus oryzae A. niger

Glucose isomerase Invertase Pullulanase Pectinase Chymosin Rennin β-Glucanase Lipase Lactase

joining these two amino acids to make aspartame involves the use of a protease working “backwards,” synthesizing a dipeptide made up of phenylalanine and aspartate. Other methods for this synthesis exist, but enzymes are favored because they preserve the correct optical properties of aspartame. This is important because L-aspartame is sweet, but D-aspartame is bitter! Theoretically, enzymes could be obtained from animals, plants, or microbes. Indeed, food-grade enzymes are obtained from all of these sources; plant-derived proteases, for example, are preferred over other sources for the purposes of meat tenderizing. However, for most other enzymatic processes, microbial enzymes are preferred, primarily because they are easier and cheaper to obtain. One reason for this is that many bacteria and fungi have the natural ability to secrete enzymes into their environment. This is an important trait for microbes that rely on the digestion of organic matter as a source of energy, carbon, and nutrients. When microbes secrete valuable enzymes, it is relatively easy to separate the enzymes from the cells — simple centrifugation may be sufficient. Another reason for relying on microbes is that several microbes, including the bacteria Bacillus subtilis and Bacillus licheniformis and the fungus Aspergillus niger, have a long history of safe use in the food industry (see Table 7.2 for other important enzyme producers). These microbes are also potent producers of a broad range of useful hydrolytic enzymes. This allows companies to supply enzymes as mixtures; for example, a company might market a product labeled as “proteases” that actually contains a number of different proteases, often produced by the same microbe. Mixtures of related enzymes are often useful, because many applications of enzymes are founded on empirical discoveries. Often, the precise mechanism of enzyme action is unknown, making it difficult to assign specific function to specific enzymes. For this reason, companies prefer to market enzymes as “categories” (e.g., proteases) rather than marketing specific enzymes.

Cell Culture and Food

215

Because the microbes used for enzyme production must be safe for food production, it is often possible to market mixtures of enzymes that have not been highly purified (in other words, purification to remove toxic compounds is unnecessary). This cuts the cost of enzyme production without compromising quality or safety. The worldwide market for industrial enzymes is around $1 billion; at least 400 companies make enzymes, although 12 companies dominate the industry. Most of the enzymes produced by these companies derive from microbes and are used by brewers, wine makers, starch processors, dairies, and many other food and beverage processors. Many nonfood applications of enzymes also exist, such as the use of proteases and lipases in laundry detergents. We will now examine the major uses of microbial enzymes in the food industry.

B. AMYLASES 1. Starch Processing Amylases are arguably the most important enzymes in the food industry. Endogenous amylases in barley are crucial in malting of barley and other cereals (see this chapter, Section II, B and C) and are added to flour to improve baking characteristics. They are the cornerstone of the processing of inexpensive starch to produce a range of value-added products that are used as food ingredients (e.g., glucose). Starch is a remarkable and useful polymer. It can be transformed into a number of products with many different uses. Some of these transformed products are useful because of their physical properties — for example, maltodextrins are important components of many confections because they give food a chewy texture and are a cheap and bland-tasting filler. Other products of starch processing are valuable sweeteners; the main source of fructose, which is sweeter than sucrose, is from the enzymatic conversion of glucose (derived from starch). Starch is the most common compound used by plants to store carbohydrates. Plant cells often contain specialized organelles known as amyloplasts that are filled with starch grains. Roots and seeds often have a high starch content, as do corms, tubers, bulbs, and fruits. Many of the food crops that humanity is highly dependent on, such as rice, wheat, corn, and potatoes, have high starch content. These are all heavy-yielding plants; consequently, there is an abundant and cheap supply of starch in most regions of the world. Combined with the development of systems for the mass production of inexpensive starch-modifying enzymes, this has led to the establishment of a large and important industry. Starch is composed of linear chains of α-1,4-linked glucose monomers (Figure 7.15). In amylose, the individual chains are unbranched; however, in amylopectin, branches occur through the formation of α-1,6 bonds. Amylopectin is more insoluble in water than amylose and is generally more difficult for microbes to fully degrade. The length of each linear chain varies widely but is often in the range of 1000 residues. A number of enzymes are involved in starch digestion and processing. Some of the following enzymes take part in starch digestion, whereas glucose isomerase and cyclodextrin glycosyltransferases drive the conversion of glucose into other useful compounds:

216

Introduction to Food Biotechnology

A. Amylose CH2OH

CH2OH O

OH

O O

OH

CH2OH O

OH

OH

CH2OH

OH

O OH

O O

OH OH

OH

OH

B. Amylopectin

CH2OH

CH2OH O

OH

O

OH

CH2OH O

O

OH

OH

OH

CH2OH

OH

CH2OH O

O

OH

OH

O α-1,6 link

OH

CH2OH O

OH

OH

O

CH2

O O

OH

OH

O OH

α-1,4 link

O

OH OH OH

C. Sites of enzyme attack Pullulanase

β-amylase

α-amylase

FIGURE 7.15 Structure of amylose (A) and amylopectin (B). Sites of enzymatic digestion of starch are also shown (C), with each circle representing a glucose monomer.

• α-Amylases attack α-1,4 links within starch. These enzymes increase the solubility of starch because they break the starch into smaller pieces (see Figure 7.15c). • β-Amylases break α-1,4 links at the nonreducing ends of starch molecules, releasing maltose (see Figure 7.15c). • Pullulanases are considered to be debranching enzymes, because they attack α-1,6 links that join linear starch chains (see Figure 7.15c).

Cell Culture and Food

217

• Glucoamylases release glucose monomers from the nonreducing ends of the starch chains. • Glucose isomerases convert glucose to fructose. • Cyclodextrin glycosyltransferases are used to convert dextrins to cyclic compounds. These compounds are not currently widely used in the food industry, but they may be in the future. They form inclusion complexes that can enclose “guest” molecules (this is an example of encapsulation); this alters properties of the guest molecule and may enhance stability or activity of enzymes or flavor compounds. These enzymes have distinct uses in starch processing. The nature of the desired end product of the process dictates whether each enzyme is used. Because of the diversity of starch-derived products, it is useful to have a system that allows rapid comparison of compounds. The dextrose equivalent (DE) system compares each compound to dextrose (glucose). If a starch molecule is completely hydrolyzed into glucose monomers, this is considered to be 100% dextrose conversion, or a product with a DE of 100. The range of DE is therefore between 0 and 100; the DE value is a good indicator of the physical nature and potential uses of a modified starch product. Unmodified starch has a DE of 0, or close to 0. It is very similar to the starch found in the source plant. It has many industrial nonfood applications and is used extensively in the food industry as a thickener (e.g., corn starch). Maltodextrins have a DE of less than 20. Like unmodified starch, they are not sweet and have a bland taste. They have a myriad of uses within the food industry. The addition of maltodextrins to a food makes the food less hygroscopic and often gives it a chewy texture. The food is also less susceptible to drying out, and if it is frozen (e.g., ice cream), ice crystal formation is inhibited. Corn syrup solids are more highly modified than maltodextrins and have a DE of more than 20. They are added to food to protect it from oxidation and to encapsulate flavors. They also affect the physical properties of food and increase its ability to maintain moisture content. Unlike maltodextrins, however, they are moderately sweet. Maltodextrins and corn syrup solids consist of a mixture of dextrins, glucose, and maltose. As the DE increases, the proportion of dextrins decreases and the proportion of maltose and glucose increases. Glucose (corn) syrup is highly modified starch and has a DE of close to 100. It is extensively used as an inexpensive sweetener in foods. It is also a popular additive to low-calorie foods because it can substitute for fats in certain processed foods. Maltose-rich syrups are similar but consist of high levels of maltose and little free glucose. These syrups are moderately sweet and have the typical “malt” flavor that is desirable in breakfast cereals, for example. Finally, high-fructose syrups are popular sweeteners in many processed foods and are made by enzymatic conversion of glucose syrups. Fructose is twice as sweet as sucrose. A range of products can be obtained from starch through the judicious use of microbial enzymes. This requires tight control over the extent and pattern of hydrolysis. In the past, this was done by chemical hydrolysis (e.g., hydrochloric acid). Enzymes are preferred because they give the processor greater control over the

218

Introduction to Food Biotechnology

process and do not result in large salt residues, as does chemical hydrolysis. A typical enzymatic starch degrading operation consists of the following steps: 1. Liquefaction. The starch is heated to 95°C and α-amylase from B. licheniformis is added. This heat-stable enzyme solubilizes the starch and releases it from starch grains. After treatment with α-amylase, the starch has a DE of 10 to 12. 2. Saccharification. Glucoamylase from A. niger generates glucose from nonreducing ends of the starch fragments. Pullulanase from Bacillus or Klebsiella is also added to break α-1,6 links in the starch. To avoid caramelization, the temperature is reduced to 60°C. The extent of saccharification can be controlled by varying the time that the starch is exposed to glucoamylase. Thus, maltodextrins, corn syrup solids, and glucose syrups are obtained through progressively longer saccharification treatment. If a high-maltose syrup is desired, then plant-derived β-amylase is used instead of glucoamylase. 3. Purification. This nonenzymatic step decolorizes the product. Typically, activated charcoal is added, then filtered out, and the filtrate is heated and evaporated to concentrate the solution into syrup. Activated charcoal absorbs many compounds, including those responsible for the brownish color of the glucose syrup. 4. Isomerization. This step is taken only if a high-fructose syrup is desired. Immobilized systems using glucose isomerase from Streptomyces spp. are popular. Usually, the enzyme is linked to a solid support, which is packed into a column. Using enzymes currently available, only 50% of a batch of glucose can be converted into fructose. We have not yet discovered an enzyme whose natural function is to catalyze the conversion of glucose to fructose in one step. Virtually all cells are capable of converting glucose to fructose, but three enzymes are required. It is much more desirable to use one enzyme in an industrial operation, so an enzyme is used that is capable of converting glucose to fructose in one step even though this is not the natural function of this enzyme. This enzyme — xylose isomerase — has a lower affinity for glucose than its normal substrate (xylose), so it is not possible to convert much more than 50% of the glucose to fructose. Fortunately, it is relatively easy to separate fructose from glucose through differential crystallization. This process is very important to the food industry — greater than 100 billion tons per year of fructose are used as sweeteners. 2. Amylases and Baking α-Amylases are added to flour to improve baking properties. Flour naturally contains amylases derived from wheat grains, but they are not always present in sufficient amounts. Addition of α-amylases improves the quality of bread and other baked goods through:

Cell Culture and Food

219

• Decreased dough viscosity, leading to easier dough handling, and perhaps contributing to increased bread volume. • Increased dextrinization during heat treatment, resulting in improved color and flavor. • Increased supply of fermentable sugars, leading to increased bread volume via increased production of CO2 by yeasts. • Improved shelf life, perhaps through the formation of low-molecularweight branched starch fragments that interfere with starch recrystallization. Recrystallization results in firmer crumb structure and an undesirable texture (hard and excessively chewy) — generally referred to as “bread staling.” The molecular mechanisms behind these improvements in bread quality are unclear but are probably related to the partial digestion of amylose and the generation of dextrins. There is currently great interest in the potential of other enzymes to improve the quality of baked goods. Hemicellulases, which release soluble carbohydrates from hemicellulose, one of the insoluble components of the plant cell wall, have received particular attention. Their use could increase the soluble fiber content of dough, resulting in improved functionality (in the health sense) of baked goods. Use of these enzymes may also improve intestinal absorption of nutrients derived from baked goods.

C. LIPASES Lipases are increasingly popular and important contributors to food processing. Two major applications of lipases are: (1) specific tailoring of lipids to produce desired fatty acid structure, and (2) generation of volatile fatty acids to improve the flavor of butter, cheese, and various other foods. The first application makes use of the synthetic capabilities of lipases, and the second makes use of the hydrolytic properties of these enzymes. Lipid tailoring is achieved by trans-esterification (Figure 7.16). Recall that simple lipids are composed of glycerol with ester linkages to three fatty acids. If a simple lipid is exposed to an additional (desired) fatty acid and lipase, the enzyme will catalyze replacement of one or more of the fatty acids by the more desirable fatty acid. Some lipases catalyze this reaction at a specific point in the glycerol molecule (e.g., a 2-specific lipase exchanges fatty acids only at the second carbon of glycerol). Other lipases (e.g., those produced by Candida rugosa, a GRAS organism) are nonspecific and can generate several different lipids. Lipases from C. rugosa are particularly popular because they can exchange a wide range of fatty acids of different lengths and degree of unsaturation. Thus, trans-esterification provides precise modification of lipid structure. This is sometimes used to produce specific types of detergents and biosurfactants. It also has great potential to increase the health-promoting aspects of vegetable oils, because it can be used to increase the content of fatty acids such as linolenic acid, which are generally beneficial to human health. This technology can also be used to modify melting points, solubility, and other physical properties of edible oils.

220

Introduction to Food Biotechnology O O O O

Initial Lipid

O

+

O O O

Desired Fatty acid HO + Lipase

O O O O O O O

FIGURE 7.16 The use of lipases to drive trans-esterification. One of the fatty acid chains is replaced by a chosen fatty acid.

To “convince” lipases to function as synthetic enzymes, one must decrease the water content to low levels (often 500,000 tons), particularly in the beverage industry, partly because of its long history of safe use in foods. It is also of considerable historical interest, because commercial citric acid production began in 1923 and was one of the first successful applications of biotechnology. Citric acid is used primarily as an acidulant to lower the pH of a food or beverage. Although increased acidity leads to inhibition of most bacteria, citric acid is not normally used as an antimicrobial (acetic acid and lactic acid are more effective). However, citric acid has a number of other properties that make it attractive as an acidulant. It imparts a pleasant sour taste to beverages (e.g., soft drinks) and candies and it enhances other flavors, especially those present in fruits and vegetables. Citric acid also functions as a metal chelator, and it is used to inhibit off flavor development in fresh and frozen fish. Dimethylamine, produced by endogenous fish

Cell Culture and Food

227

enzymes, leads to an unpleasant “fishy” flavor. Dimethylamine production is inhibited by citric acid through chelation of iron and copper ions. Finally, citric acid is commonly used as an antioxidant, particularly in combination with ascorbic acid. This combination is often used to prevent browning in fruits and vegetables. In this context, chelation of metal ions by citric acid inhibits polyphenol oxidase, a cause of browning.

B. PRODUCTION

OF

CITRIC ACID

BY

ASPERGILLUS

NIGER

A number of fungi can be used to produce citric acid on an industrial scale, but A. niger is currently the most popular. This fungus does not normally produce large amounts of citric acid, but overproduction is possible because of the development of: (1) overproducing strains of A. niger (this has mainly been accomplished by extensive screening of isolates), and (2) growth conditions that lead to citric acid overproduction. To overproduce citric acid, assuming that an efficient strain has been found, three conditions must be met. First, because A. niger is a strict aerobe, it must have a sufficient supply of oxygen. Second, the fungus must be deficient in manganese. Third, high levels of glucose must be present. We do not have a complete understanding of the importance of manganese deficiency and high glucose, but details are starting to become clear. Citrate is not the end product of a fermentation or other pathway in A. niger. Instead, it is an intermediate in an important cyclic pathway — the TCA cycle. Why would a fungus produce large amounts of such an intermediate compound? The answer appears to lie in the regulatory function of two key enzymes (Figure 7.20): phosphofructokinase (PFK), a glycolytic enzyme, and α-ketoglutanate dehydrogenase (α-KDH), an enzyme in the TCA cycle. Citric acid accumulation is normally prevented because citric acid allosterically inhibits PFK. Inhibition at this stage results in little flow of glucose through glycolysis and to the TCA cycle. However, manganese deficiency results in the loss of this feedback inhibition. This is probably linked to amino acid metabolism and rates of protein turnover. Manganesedeficient hyphae have increased levels of protein turnover, leading to a requirement for increased intracellular levels of NH4+. Increased NH4+ appears to be crucial to relief of allosteric inhibition of PFK by manganese deficiency. α-KDH is the key enzyme regulating catabolism of citrate in the TCA cycle. It is inhibited by oxaloacetate, one of the later intermediates in the TCA cycle. Oxaloacetate production through carbon fixation (pyruvate + CO2 → oxaloacetate) is in turn stimulated by high levels of carbohydrates in the external environment. Thus high levels of glucose or other carbohydrates stimulate oxaloacetate production, which in turn inhibits activity of α-KDH. This model has not been completely verified and may be adjusted in the future. High levels of carbohydrates also act to lower the Km of PFK, thus increasing its activity and increasing the flux of carbon through glycolysis. This carbon cannot be oxidized by the TCA cycle, so the cell gets rid of it through citric acid secretion. We are approaching a mechanistic understanding of the empirical observation that high levels of carbohydrates combined with low levels of manganese lead to

228

Introduction to Food Biotechnology Glucose PFK

Glycolysis

Pyruvate Acetyl co-A

Pyruvate + CO2

Citrate Oxaloacetate Isocitrate Malate

TCA cycle

a-ketoglutarate a -KDH

Fumarate Succinate

FIGURE 7.20 Overproduction of citric acid in Aspergillus niger. Under conditions of high glucose, high oxygen, and low manganese, citric acid is overproduced. High glucose levels result in increased conversion of pyruvate to oxaloacetate. Increased oxaloacetate levels inhibit (dotted line) α-KDH, restricting the flow of carbon through the TCA cycle. High glucose also increases the activity of phosphofructokinase (PFK), which increases the flux of carbon through glycolysis. This extra carbon cannot be oxidized by the TCA and is secreted as citric acid.

citric acid production. But what does the microorganism gain from these biochemical machinations? Is there any cellular benefit to the overproduction of citric acid under these conditions? At present, our limited understanding of the cellular ramifications of manganese deficiency makes it difficult to answer this question.

C. VITAMIN PRODUCTION

BY

MICROORGANISMS

Vitamin enrichment and fortification is an efficient way of encouraging vitamin sufficiency among consumers. Many people consume vitamin supplements to compensate for poor diets or for other health reasons, or because of conditions such as pregnancy that require increased vitamin supply. Many food processes (e.g., heat treatments) result in losses of vitamins that need to be replaced somehow. In the developing world, vitamin supply is particularly problematic because widespread poverty prevents many people from consuming a diet adequate in all vitamins. Vitamin deficiency leads to a host of health problems (e.g., increased frequency of bacterial infections). Consequently, not only is vitamin production economically important, but it affects global health considerably.

Cell Culture and Food

229

Most vitamins are made through chemical synthesis. However, a number of vitamins, including ergosterol (provitamin D), riboflavin, B12, and B13, are produced through microbial culture. Others, such as vitamin C, are produced through a combination of chemical synthesis and microbial metabolism. Microbial production of vitamins is mainly achieved through selection of natural (wild) overproducing strains, followed by the design of bioreactor conditions that optimize production. For example, two species of filamentous fungi (Eremothecium ashbyi and Ashbya gossypii) that are closely related to S. cereviseae, naturally overproduce riboflavin. These and other wild overproducers can now be used to produce riboflavin at similar costs as chemical synthesis. This has led several large vitamin companies to phase out chemical synthesis and replace it with microbial synthesis, which also has the advantage of being perceived as a “green” process that uses renewable substrates. In contrast, chemical synthesis relies on nonrenewable fossil fuels as substrates for vitamin production. One company (Roche) is replacing chemical synthesis of riboflavin with microbial synthesis using mutant strains of B. subtilis. This bacterium is not naturally an overproducer of riboflavin, but it is a GRAS organism that is already widely used in the food industry. We expect that the trend away from chemical synthesis to microbial production will spread to other vitamins, particularly if forecasted shortages of fossil fuels in the mid-20th century drive petroleum prices upward.

VIII. DEVELOPMENT OF NOVEL MICROBIAL PRODUCTS Although many screening programs have been aimed at finding strains of microbes that “do useful things,” undoubtedly there remain a multitude of useful microbes that have not been discovered. This is especially true in reference to microbial polysaccharides. There is enormous range in the structure of these compounds, and future biotechnologists will certainly exploit this variation. However, successful exploitation of a microbe (or cell culture) is not always easy, and is often impossible. To assess the potential success of a microbial venture, it is useful to ask a number of questions about the microbe and its product: 1. Is there a market for the product? Obviously, if the answer is no, then there is no reason to continue. If the answer is yes, then several other questions become pertinent. For example, it is useful to estimate the range of prices of the product that would be tolerated by the market. 2. Can the product be obtained from nonmicrobial sources? If the product can be obtained from plants or seaweeds, then it is unlikely that microbial production will be able to compete. Plants are fundamentally easier and cheaper to grow than cells or microbes. There are exceptions; for example, vanillin can be produced from large-scale cell culture at comparable costs as from soil-grown plants. 3. Does the microbe (or cell) produce large amounts of the desired product? If the answer is no, then a program for strain improvement is required.

230

Introduction to Food Biotechnology

Collect and assess a large number of isolates. Expose the best producers to mutagens and screen the resultant mutants. This strategy has been successful in dramatically increasing the production rate of many important microbial compounds. Recombinant procedures can also be applied to most microbes to increase the rate of production, but this requires substantial investment in basic research into the physiology and genetics of biosynthesis of the desired compound. Changes in the bioreactor environment can also strongly affect production and are usually discovered through careful empirical research. 4. Is the microbe (or cell) easy to grow? If the microbe requires expensive and well-defined media, then it is unlikely that production on an industrial scale will be feasible. However, if the microorganism can be grown on cheap substrates such as molasses or corn steep liquor, then the chances of success are much greater. If the microbe is difficult to grow, it may be possible to transfer the genetic material responsible for product formation to a microbe that is easier to grow. 5. Is the microbe safe? This is a crucial question. If the microbe has been granted GRAS status, it can be used in food production. However, if the microbe is not GRAS, you must expend great effort to demonstrate that it is safe. You must clearly show that the microbe is not pathogenic to humans and does not produce any toxins. Furthermore, it must not produce antibiotics, and you must demonstrate that the desired product can be purified without the introduction of harmful contaminants (e.g., organic solvents). Accumulating the evidence required to obtain GRAS status is an expensive process. For this reason, it is vastly preferable to use an organism that already has GRAS status. Consequently, it is not surprising that so many enzymes are obtained from a small number of microbes (e.g., B. subtilis, A. niger, S. cereviseae). An alternative to acquiring GRAS status is to transfer the necessary genes to a GRAS organism. However, most countries still require extensive safety testing of the resultant recombinant microbe. This is particularly true if the source of the DNA is a pathogenic, or toxin-producing, microbe. In such cases, it may be impossible to obtain permission to sell the recombinant product. Once all these questions have been answered satisfactorily, a decision can be made about the potential for the microbial process to be successfully scaled up to an industrial scale. The large number of successful microbial biotechnologies suggests that despite the onerous challenges that are encountered, the large-scale culture of microorganisms is often both desirable and feasible. Once a suitable microbe has been chosen for a particular process, the biotechnologist is unlikely to be completely satisfied. Microbial biotechnology is always governed by the bottom line; any changes to a microbe or a microbial process that increase efficiency and decrease costs will be desirable. Strain improvement is therefore an essential and continuing process. The best example of strain improvement is in the antibiotic industry, where dramatic increases in production efficiency have occurred as a result of intensive improvement programs. The techniques of

Cell Culture and Food

231

strain improvement are similar to those used in screening programs. Mutagenesis is often a highly successful strategy, but the current trend is toward exploitation of basic understanding of genetic regulation of product formation, and of interrelationships among metabolic pathways (quantitative physiology). As our understanding of microbial metabolism increases, our power to tailor metabolism to our needs will also grow.

RECOMMENDED READING 1. Glazer, A. N. and Nikaido, H., Microbial Biotechnology: Fundamentals of Applied Microbiology, Freeman and Company, New York, 1995. 2. Linko, M., Haikara, A., Ritala, A., and Penttilä, M., Recent advances in the malting and brewing industry, J. Biotechnol., 65, 85, 1998. 3. van Dam, K., Role of glucose signaling in yeast metabolism, Biotechnol. Bioeng., 52, 161, 1996. 4. Randez-Gil, F. and Sanz, P., Construction of industrial baker’s yeast strains able to assimilate maltose under catabolite repression conditions, Appl. Microbiol. Biotechnol., 42, 581, 1994. 5. Jensen, L. G., Olsen, O., Kops, O., Wolf, N., Thomsen, K. K., and von Wettstein, D., Transgenic barley expressing a protein-engineered, thermostable (1,3-1,4)-β-glucanase during germination, Proc. Natl. Acad. Sci. U.S.A., 93, 3487, 1996. 6. McCaig, R., McKee, J., Pfisterer, F. A., Hysert, D. W., Munoz, E., and Ingledew, W. M., Very high gravity brewing — laboratory and pilot plant trials, J. Am. Soc. Brew. Chem., 50, 18, 1992. 7. Suihko, M.-L., Vilpola, A., and Linko, M., Pitching rate in high gravity brewing, J. Inst. Brew., 99, 341, 1993. 8. Onnela, M.-L., Suihko, M.-L., Penttilä, M., and Keränen, S., Use of a modified alcohol dehydrogenase, ADH1, promoter in construction of diacetyl non-producing brewer’s yeast, J. Biotechnol., 49, 101, 1996. 9. Varnum, A. H., The exploitation of microorganisms in the processing of dairy products, in Exploitation of Microorganisms, Jones, D. G., Ed., Chapman & Hall, London, 1993, chap. 11. 10. Chapman, H. R. and Sharpe, M. E., Microbiology of cheese, in Dairy Microbiology, Vol. 2, The Microbiology of Milk Products, Elsevier, London, 1990, chap. 5. 11. Martin, M. C., Alonso, J. C., Suárez, J. E., and Alvarez, M. A., Generation of foodgrade recombinant lactic acid bacterium strains by site-specific recombination, Appl. Environ. Microbiol., 66, 2599, 2000. 12. Gilbert, C., Robinson, K., Le Page, W. F., and Wells, J. M., Heterologous expression of an immunogenic pneumococcal type 3 capsular polysaccharide in Lactococcus lactis, Infect. Immun., 68, 3251, 2000. 13. Hashimoto, S.-I. and Ozaki, A., Whole microbial cell processes for manufacturing amino acids, vitamins or ribonucleotides, Curr. Opinion Biotechnol., 10, 604, 1999. 14. Kumagai, H., Microbial production of amino acids in Japan, Adv. Biochem. Eng., 69, 71, 2000. 15. Whitaker, J. R., Enzymes, in Food Chemistry, 3rd ed., Fennema, O. R., Ed., Marcel Dekker, New York, 1996, chap. 7. 16. Poutanen, K., Enzymes: an important tool in the improvement of the quality of cereal foods, Trends Food Sci. Technol., 8, 300, 1997.

232

Introduction to Food Biotechnology

17. Benjamin, S. and Pandey, A., Candida rugosa lipases: molecular biology and versatility in biotechnology, Yeast, 14, 1069, 1998. 18. Lang, C. and Dörnenburg, H., Perspectives in the biological function and the technological application of polygalacturonases, Appl. Microbiol. Biotechnol., 53, 366, 2000. 19. Becker, A., Katzen, F., Pühler, A., and Ielpi, L., Xanthan gum biosynthesis and application: a biochemical/genetic perspective, Appl. Microbiol. Biotechnol., 50, 145, 1998. 20. Roller, S. and Dea, I. C. M., Biotechnology in the production and modification of biopolymers for foods, Crit. Rev. Biotechnol., 12, 261, 1992. 21. Banik, R. M., Kanari, B., and Upadhyay, S. N., Exopolysaccharide of the gellan family: prospects and potential, World J. Microbiol. Biotechnol., 16, 407, 2000. 22. Sutherland, I. W., Microbial polysaccharide products, Biotechnol. Gen. Eng. Rev., 16, 217, 1999. 23. Bigelis, R. and Tsai, S.-P., Microorganisms for organic acid production, in Food Biotechnology: Microorganisms, Hui, Y. H. and Khachatourians, G. G., Eds., Wiley, New York, 1994, chap. 6. 24. Vandamme, E. J., Production of vitamins, coenzymes and related biochemicals by biotechnological processes, J. Chem. Tech. Biotechnol., 53, 313, 1992. 25. Stahmann, K.-P., Revuelta, J. L., and Seulberger, H., Three biotechnical processes using Ashbya gossypii, Candida famata, or Bacillus subtilis compete with chemical riboflavin production, Appl. Microbiol. Biotechnol., 53, 509, 2000. 26. Parekh, S., Vinci, V. A., and Strobel, R. J., Improvement of microbial strains and fermentation processes, Appl. Microbiol. Biotechnol., 54, 287, 2000. 27. Steele, D. B. and Stowers, M. D., Techniques for selection of industrially important microorganisms, Annu. Rev. Microbiol., 45, 89, 1991.

8

Industrial Cell Culture I. SCALE-UP OF CELL CULTURE

Biotechnology often requires the growth of large amounts of microbial cells or of cells derived from plants or animals. In this chapter, the focus is primarily on the methods of and approaches to large-scale growth of bacteria and fungi, because these microbes produce virtually all food-related cell products. When growing bacteria or fungi in an experimental context, we use petri dishes or small volumes of liquid media. In these small-scale systems, nutrients diffuse to cells or, in the case of fungi growing on a petri dish, cells (hyphae) grow toward the nutrients. In a petri dish, all cells are continually exposed to a moist atmosphere made up of 20% oxygen, more than enough to fulfill the needs of a strict aerobe (Figure 8.1). Growing microbes in small volumes of liquid media is also straightforward; mild agitation of the cultures or growth in shallow flasks is sufficient to ensure adequate oxygen supply to the cells. This type of small-scale culture system is well suited to the experimental manipulation of cells. Much of our understanding of cellular metabolism and behavior has been obtained through small-scale cell culture. However, when the purpose of cell growth is to collect a useful product, and then to sell it, small-scale culture is no longer feasible. The operation must be scaled-up, so that the product can be efficiently collected with a minimum of labor materials. A scaled-up culture system is called a bioreactor. Bioreactors have traditionally been referred to as fermenters. This term is still widely used, but many biotechnologists prefer the term bioreactor, because most applications of large-scale microbial culture rely on aerobic metabolism, not fermentation. The size of bioreactors for growing plant and animal cells may be relatively modest, particularly in the case of animal cells. It is difficult to grow animal cells in volumes greater than 100 L because of their fragility. Nevertheless, some companies successfully grow animal cells (e.g., hybridomas to produce monoclonal antibodies) in volumes up to 10,000 L. Bacterial and fungal cells, in contrast, are frequently grown in 10,000 to 100,000-L bioreactors. Scale-up of a culture system from a 2-L flask to a 1000-hL bioreactor will invariably result in problems that did not exist in the 2-L flask. Agitation of the bioreactor is not an option. One must find other methods for aerating the culture medium when growing aerobic cells. Diffusion will no longer suffice to supply nutrients throughout the bioreactor. Both of these problems (nutrient and oxygen supply) can be tackled together, through aeration (bubbling oxygen or air into the bioreactor) or agitation by mechanically stirring the liquid medium.

233

234

Introduction to Food Biotechnology O2

A. Petri Dish

B. Tube

C. Flask

FIGURE 8.1 Oxygen supply to small-scale cell cultures. In petri plates (A), oxygen diffuses through the space between the top and bottom covers of the plate. In small tubes (B), agitation results in mixing of the liquid medium (solid arrows). Oxgen diffuses from the air column above the liquid (dashed arrows). A small flask (C) is often used to grow animal cells. The flask is laid on its side with a thin layer of liquid growth medium. Oxygen diffuses through the loosely attached cap (not shown) and into the flask. Oxygen then diffuses into the liquid.

The temperature within the bioreactor may quickly rise to inhibitory levels, resulting in the need for an efficient cooling system. The biotechnologist may find, to his/her dismay, that after inoculating the bioreactor with the desired organism, appreciable growth in the bioreactor does not occur for several weeks. It may be necessary to prepare large amounts of inoculum in a smaller bioreactor, which can then be used to seed the larger bioreactor, to avoid long delays due to cell division. As a final insult, the culture may grow well in the bioreactor but fail to make appreciable amounts of the desired product. Microbiologists frequently observe changes in cellular morphology and metabolism upon growth in large bioreactors that may lead to decreased product formation. An even worse scenario is possible — sometimes scale-up leads to frequent contamination by undesirable microbes. If undetected, this could cause major problems for consumers and for the company; contamination can lead to off odors or undesirable flavors, and, if the contaminant is pathogenic or toxigenic, consumers may suffer greatly. For these reasons, scale-up typically occurs through a gradual increase of bioreactor size. Initially, the cell may be grown in a 100-L container, to find out if this increase in culture volume changes growth rates or rates of product formation. The operation will then be scaled up to the pilot plant stage, which may have a volume of 1,000 to 10,000 L. The pilot plant will also incorporate pilot versions of collection and purification systems. When all problems have been resolved in the pilot plant, the company will shift to the final bioreactor volume and perform further tests to ensure that cell growth and product formation occur in the predicted manner. Scaling up an operation also forces one to carefully consider the growth medium. Is it expensive? Do media exist that offer equivalent growth rates at a fraction of the cost? Are some of the components of the medium unnecessary? Such questions quickly become important as the culture volume increases, and they can make the

Industrial Cell Culture

235

TABLE 8.1 Oxygen Requirements of Cells Grown in Large-Scale Systems Organism

Product

Oxygen Requirements

Aspergillus oryzae Bacillus subtilis Lactobacillus spp. Saccharomyces cereviseae S. cereviseae Corynebacterium glutamicum. Zymomonas mobilis

Various enzymes Various enzymes Starter cultures Ethanol Biomass (baking) Glutamate Ethanol

Aerobic Aerobic Oxygen not required Anaerobica Aerobic Low concentration of oxygen Anaerobic

a

S. cereviseae requires oxygen during the initial phase of ethanol fermentation.

difference between a process that is financially feasible and one that would result in a product that is too expensive for its intended market. It is equally important to accurately assess the marketplace. Will the targeted consumers desire the product, and if so, can it be sold at an affordable price? This is particularly true in the food industry. Most of the microbial products destined for the food industry are food additives. Most have a range of competitors that are produced through alternative means (e.g., chemical synthesis or derived from plants or animals). Consequently, products of cell culture must offer superior performance or must be competitively priced. Economic factors are particularly important for young biotechnological companies. Delays in production caused by scale-up problems can lead to severe cash flow problems, particularly in the current environment where venture capital is difficult to acquire. We will now discuss aspects of the bioreactor environment that have the most important effects on bioreactor design and function. They are oxygen supply, heat removal, and nutrient supply. Since the original development of industrial-scale microbial culture in the 1920s, scale-up problems associated with these factors have largely been resolved through collaborative efforts of microbiologists, biochemists, and biochemical engineers. In the next section, the major aim is to develop a basic understanding of the approaches taken to ensure that cells in a bioreactor are exposed to an optimal environment for growth and product formation.

II. ENVIRONMENTAL FACTORS A. OXYGEN One of the basic principles of microbiology is that microbes exhibit great variation in their relationship to oxygen (Table 8.1). Strict aerobes require oxygen, whereas others, known as strict anaerobes, are killed by oxygen. The industrially important lactic acid bacteria (LAB) are aerotolerant, meaning that they do not use oxygen but can tolerate its presence. The amount of oxygen an organism requires affects bioreactor design. Anaerobic organisms are much easier to grow in large volumes

236

Introduction to Food Biotechnology

than aerobes, because of the low solubility of oxygen in water (8 mg/L at 30°C). Rapidly growing aerobic organisms consume large amounts of oxygen. For example, an actively respiring culture of Saccharomyces cereviseae consumes 6 g O2/(L · h). Oxygen is rapidly depleted, resulting in anaerobic zones around growing cells. Oxygen depletion is relieved by diffusion of oxygen from the surrounding liquid, but this is a slow process. Many factors affect the diffusion rate of a soluble compound; one of the most important is the strength of the concentration gradient. Diffusion is rapid from a region with a high concentration to a region with a low concentration. With oxygen, concentration gradients are relatively small, because of the low solubility of oxygen in water; thus, diffusion of oxygen is relatively slow. Diffusion is also much slower if the distance of diffusion is large. Therefore, the biotechnologist who wants to grow large amounts of aerobic cells is faced with two problems: (1) when large amounts of cells are actively metabolizing, the oxygen concentration around the cells quickly falls to low levels; and (2) as the vessel size increases, the distance to the source of oxygen (the surface of the liquid) also increases. The interplay of these factors results in severe stress to aerobic organisms; the lack of oxygen will inhibit or kill them, and the desired product will not be produced. Why do cells require oxygen? Remember that virtually all eukaryotic organisms produce adenosine triphosphate (ATP) through respiration of carbohydrates. Oxygen is the terminal electron acceptor in the electron transport chain that drives ATP formation, and low levels of oxygen severely inhibit respiration in most eukaryotes. Consequently, the supply of oxygen to plant and animal cells is crucial to their growth and metabolism. Eukaryotic fungi are often more flexible in their oxygen requirements. For example, the yeast S. cereviseae is able to survive and grow under conditions of low oxygen concentration because of its ability to ferment carbohydrates such as glucose or sucrose into ethanol. Because fermentation yields less ATP than aerobic respiration, yeasts in anaerobic environments grow more slowly than under aerobic conditions. When ethanol is the desired end product, as is the case in the production of beer or wine, this slow growth is not important, because large amounts of ethanol can be produced by slowly growing cells. However, if we want to produce large amounts of biomass of S. cereviseae, we must ensure that the cultures are well supplied with oxygen. Thus, different bioreactor strategies are required by companies that produce and sell baker’s yeast (viable cells of S. cereviseae) as compared to companies that produce ethanol or ferment beer or wine. Not all yeasts are able to grow under anaerobic conditions. For example, Kluyveromyces is used extensively to produce enzymes such as lactase and recombinant proteins such as chymosin. It is a strict aerobe and thus must be well supplied with oxygen. Many prokaryotes are capable of anaerobic growth, but many are strict aerobes. Xanthomonas campestris, for example, is grown in large quantities to produce xanthan gum and is a strict aerobe. Facultatively anaerobic prokaryotes such as Escherichia coli are usually grown aerobically in biotechnological processes, because more rapid growth typically results in better rates of production of the desired compound.

Industrial Cell Culture

237

Bioreactor wall

Impeller blade Air bubble Impeller shaft

Sparger

FIGURE 8.2 Use of a sparger to inject air bubbles into a bioreactor and an impeller to mix the bubbles throughout the bioreactor. As the impeller shaft rotates (arrow), the blades agitate the liquid medium.

Some microbes require oxygen, but at low levels. Corynebacterium glutamicum produces large amounts of glutamate, an amino acid that is an important flavor enhancer of food. This bacterium is capable of growth under aerobic conditions, but, because of a peculiarity in its carbohydrate catabolism (see Chapter 7, Section IV.D), it overproduces glutamate only when oxygen is present at a low concentration. This bacterium is not a micro-aerophile (organism that grows best at low oxygen levels). It is able to grow well under aerobic conditions, but little glutamate is made when oxygen is abundant. Prior to the 1940s, most large-scale cell culture involved the growth of facultative or strict anaerobes (citric acid production by the aerobe Aspergillus niger is the major exception). However, World War II created an enormous demand for penicillin, the newly discovered antibiotic. Efforts to grow large quantities of Penicillium notatum were initially unsuccessful because of its oxygen requirements. The development of surface culture to grow P. notatum solved this problem. The fungus was inoculated onto solid substrates in trays; the fungus grew on the surface of the trays and oxygen was supplied directly from the atmosphere. This method was used until technology was developed that allowed growth of aerobic cells in liquid culture. The major problem with surface culture is that a large surface area is required. Such extensive surfaces are expensive to construct, and inoculation of each tray is time consuming and expensive. Furthermore, these systems are vulnerable to contamination by airborne propagules of microbes. Most aerobic systems are based on liquid culture, although solid support systems are increasing in popularity. The problem of oxygen supply in a liquid-filled bioreactor is solved by the use of impellers (Figure 8.2). Impellers consist of a rotating shaft with blades that extend into the bioreactor; rotation of the impeller mixes liquids within the bioreactor. In most cases, air is also bubbled into the

238

Introduction to Food Biotechnology

bioreactor. The impeller distributes the air bubbles throughout the bioreactor, and oxygen diffuses from the bubbles into the liquid and into cells. This combination of aeration and impeller is typical of stirred-tank bioreactors (STBs), the most common type of bioreactor. In some cases, injecting air or oxygen into the medium provides enough agitation to distribute bubbles throughout the bioreactor; if this is sufficient, then an impeller is unnecessary. Agitation can also mix nutrients throughout the bioreactor, which can help prevent the occurrence of zones that are severely depleted of nutrients due to cell growth. Impellers aerate large vessels quite well; however, they also generate shear stress. The concept of shear stress can be visualized by imagining the flow of liquid within a bioreactor. In the direction perpendicular to the flow, a gradient in velocity exists between moving and stationary liquid. This velocity gradient constitutes shear stress. The flow velocity of the liquid is known as shear strain — the moving impeller exerts pressure on the liquid, and this causing strain, or deformation. Liquids cannot be compressed, so deformation leads to flow. As the impeller rotates more quickly, shear strain and shear stress increases. The concept of shear stress is relevant to cell metabolism and viability; some cells (e.g., animal cells) are very sensitive to shear stress, whereas others (e.g., most bacteria) are relatively insensitive to shear stress. Filamentous fungi tolerate shear stress well, but they change in morphology — they typically grow as spherical pellets within STBs. Product formation rates are often related to pellet diameter, so it is necessary to maintain a level of shear stress that leads to optimal pellet size. Because of the undesirable effects of agitation on cells, and because of the expenses involved in agitation and aeration, it is important to be able to quantify the amount of aeration and agitation required to satisfy cellular oxygen requirements. The following formula is often used to model oxygen supply (OTR, oxygen transfer rate) to a bioreactor: OTR = kLa (c* – cL)

(8.1)

a is the specific interfacial area, the area per unit of culture liquid that participates in oxygen transfer. This can be visualized by imagining a stream of air bubbles injected into a bioreactor. a is correlated with the rate of air flow into the bioreactor, the volume of air injected, and the size of bubbles in the culture liquid. kL is the mass transfer coefficient. c* represents the oxygen concentration at the gas–liquid interface (in a bubble of air that is injected into the bioreactor, for example) and cL is the concentration of oxygen within the culture liquid. Therefore, (c* – cL) is a measure of the gradient in oxygen between oxygen-saturated water and the liquid of the bioreactor. It is important to determine the lower (critical) limit of cL that does not disrupt cell growth or product formation. In most cases cells do not require oxygen-saturated water; the level required will depend on the cells’ oxygen consumption rate and on other limiting factors within the bioreactor. For example, cellular respiration may be limited by substrate deficiency rather than oxygen supply. In this case, increasing the oxygen transfer rate would not result in increased growth. Some microbes (e.g., Penicillium spp.) have relatively high critical oxygen levels; when culturing these organisms, one must pay careful attention to oxygen supply.

Industrial Cell Culture

239

kLa, the volumetric mass transfer coefficient, is a crucial parameter to Equation 8.1. If kLa is too small, the oxygen transfer rate will be insufficient to supply oxygen to cells in the bioreactor; if it is too large, energy will be wasted through excessive agitation and aeration. kLa is affected most strongly by the power applied to rotate the impeller and by the aeration rate. Fortunately, it is possible to calculate the amount of aeration that is required, as long as the following information is available: • The cell’s growth rate (µ x). • The size of the initial inoculum (cx). • The amount of oxygen needed per kilogram of biomass. This is the yield coefficient (YO/X). These parameters can be used to estimate the oxygen demand of a cell culture. Oxygen demand = µ xcx/YO/X

(8.2)

For a steady state to be achieved (i.e., the oxygen level in the bioreactor remains stable and supplies just enough oxygen for the cells), oxygen supply must equal oxygen demand. We can state this mathematically by combining Equations 8.1 and 8.2: kLa (c* – cL) = µ xcx/YO/X

(8.3)

The major task facing the engineer at this point is to determine the type of reactor, operating conditions (e.g., power input to the impeller), and materials that will provide sufficient oxygen supply. Correlations between these bioreactor parameters and kLa are used to determine the required bioreactor design. Other aspects of bioreactor function must also be satisfied by the impeller and aeration combination. Mass transfer of nutrients must occur fast enough to feed growing cells (this is usually only a problem with micronutrients), and there must be enough mixing to distribute heat evenly throughout the bioreactor. Cooling systems usually consist of refrigeration or other mechanisms that operate outside of the bioreactor; consequently, heat can be removed only from the outer regions of liquid within a bioreactor. The mixing efficiency is quantified by the mixing time. This is measured by adding a substance to the bioreactor and then recording the time required to achieve uniform distribution of the substance within the bioreactor. Mixing time increases as bioreactor volume increases. For example, 29 s may suffice for an 1,800-L bioreactor, whereas 140 s are required for a 120,000-L fermenter. This requirement may lead to regions within the bioreactor that have low concentration of oxygen or nutrients. Mixing efficiency also decreases as viscosity increases. This is a problem when growing filamentous fungi, because the fungal colonies increase the viscosity of the culture liquid. This problem has an extra element of complexity because filamentous fungi exhibit non-Newtonian fluid properties (non-Newtonian fluids have a complex relationship between viscosity, shear stress, and shear strain). This makes it more difficult to predict optimal levels of gas supply and impeller operation.

240

Introduction to Food Biotechnology

Draft tube

FIGURE 8.3 An air-lift bioreactor with an inner draft tube to encourage mixing throughout the column. Air bubbles up from the bottom of the bioreactor, and, as it rises to the top of the tower, it agitates the culture medium.

The geometry of the bioreactor also affects oxygen supply. It is usually desirable to maintain as much geometric similarity as possible among the different sizes of bioreactor used in the scale-up process. Theoretical and experimental study of bioreactor geometry has led to a number of alternatives. One of these is the air-lift bioreactor, which is a cylindrical tower. Oxygen supply and mixing are achieved through bubbling air at the bottom of the tower. As the bubbles rise, they mix and aerate the culture liquid. This type of design is popular in large bioreactors (>100,000 L) where the use of impellers is impractical. Circulation within an air-lift bioreactor can be improved by addition of a draft tube, an inner cylinder that directs circulation of the culture liquid. This results in more efficient mixing of large bioreactors (Figure 8.3).

B.

PH

Microorganisms are sensitive to changes in pH. Bacteria are particularly sensitive to acidic conditions, and few bacteria (the LAB being notable exceptions) will grow at pH 8.

Industrial Cell Culture

241

The strong effects of pH on growth are important to biotechnologists, because cellular metabolism tends to cause changes in the pH of the external environment. The clearest example of this can be found in many fermentative bacteria; LAB such as Lactococcus lactis are incapable of respiration and rely on the fermentation of carbohydrates into lactic acid as a source of energy. Thus, large amounts of lactic acid are secreted by LAB, often leading to decreased pH in their external environment. Other types of cells can decrease the pH of their growth medium. For example, mammalian cells must be grown in well-buffered solutions; growth and metabolism results in the production of acidic end products that rapidly kill mammalian cells if the medium is unbuffered. The extreme sensitivity of these cells to pH changes is not surprising, because of the narrow range of pH that exists in blood (pH 7.36 to 7.42). Mammalian cells in their natural environment are constantly protected from changes in pH by several distinct buffer systems. Most culture media contain a number of compounds with buffering ability that help to stabilize the pH. However, it is advisable to monitor pH, either by taking samples periodically or by using a continuous pH detector with a probe within the bioreactor. Such systems are usually directly connected to pumps that are activated when fluctuations occur. When the pH decreases below a threshold level, a pump is activated that releases a base into the bioreactor; if the pH increases, another pump is activated that releases an acid. This type of system can effectively regulate pH in large bioreactors.

C. TEMPERATURE Temperature has strong effects on the growth rate of cells. Because waste heat is released from growing cells, bioreactors usually have cooling systems that allow regulation of temperature, preferably a temperature that allows optimal growth rates. This is an important component of bioreactor control, because most cells have a narrow range of temperature tolerance. Another reason for incorporating efficient cooling into bioreactor design is that steam is frequently used to sterilize bioreactors and culture liquids before inoculation. After such a treatment, rapid cooling to avoid extensive down time is desirable. Cooling is usually achieved by one of three mechanisms (Figure 8.4): (1) a water-filled jacket that surrounds the bioreactor; (2) coils of water pipes that surround the bioreactor or are enclosed within the bioreactor; or (3) flow of culture fluid through a heat exchanger. Heat exchangers are also required for options (1) and (2), to cool water flowing from the jacket or coils. Cooling jackets are usually restricted to laboratory-scale bioreactors (